{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 3.197314256024939, "eval_steps": 500, "global_step": 160000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0001998321410015587, "grad_norm": 4.853196620941162, "learning_rate": 1.125e-07, "loss": 0.5411, "step": 10 }, { "epoch": 0.0003996642820031174, "grad_norm": 6.491400718688965, "learning_rate": 2.375e-07, "loss": 0.5705, "step": 20 }, { "epoch": 0.000599496423004676, "grad_norm": 5.560707092285156, "learning_rate": 3.625e-07, "loss": 0.5533, "step": 30 }, { "epoch": 0.0007993285640062348, "grad_norm": 4.7198967933654785, "learning_rate": 4.875e-07, "loss": 0.5358, "step": 40 }, { "epoch": 0.0009991607050077935, "grad_norm": 5.183598518371582, "learning_rate": 6.125000000000001e-07, "loss": 0.5066, "step": 50 }, { "epoch": 0.001198992846009352, "grad_norm": 5.400509834289551, "learning_rate": 7.375e-07, "loss": 0.5301, "step": 60 }, { "epoch": 0.0013988249870109108, "grad_norm": 3.7552599906921387, "learning_rate": 8.625000000000001e-07, "loss": 0.4308, "step": 70 }, { "epoch": 0.0015986571280124696, "grad_norm": 2.8051106929779053, "learning_rate": 9.875e-07, "loss": 0.3835, "step": 80 }, { "epoch": 0.0017984892690140282, "grad_norm": 2.91430401802063, "learning_rate": 1.1125e-06, "loss": 0.3073, "step": 90 }, { "epoch": 0.001998321410015587, "grad_norm": 1.6047766208648682, "learning_rate": 1.2375000000000002e-06, "loss": 0.2899, "step": 100 }, { "epoch": 0.0021981535510171455, "grad_norm": 0.9257712960243225, "learning_rate": 1.3625e-06, "loss": 0.2249, "step": 110 }, { "epoch": 0.002397985692018704, "grad_norm": 0.8322685360908508, "learning_rate": 1.4875e-06, "loss": 0.2437, "step": 120 }, { "epoch": 0.002597817833020263, "grad_norm": 0.5329452753067017, "learning_rate": 1.6125e-06, "loss": 0.1558, "step": 130 }, { "epoch": 0.0027976499740218217, "grad_norm": 0.40953269600868225, "learning_rate": 1.7375000000000003e-06, "loss": 0.1676, "step": 140 }, { "epoch": 0.0029974821150233802, "grad_norm": 0.4813768267631531, "learning_rate": 1.8625e-06, "loss": 0.1849, "step": 150 }, { "epoch": 0.0031973142560249392, "grad_norm": 0.6746397614479065, "learning_rate": 1.9875e-06, "loss": 0.1584, "step": 160 }, { "epoch": 0.003397146397026498, "grad_norm": 0.3994137942790985, "learning_rate": 2.1125e-06, "loss": 0.1478, "step": 170 }, { "epoch": 0.0035969785380280564, "grad_norm": 0.3565943241119385, "learning_rate": 2.2375e-06, "loss": 0.135, "step": 180 }, { "epoch": 0.003796810679029615, "grad_norm": 0.37551775574684143, "learning_rate": 2.3625000000000003e-06, "loss": 0.1572, "step": 190 }, { "epoch": 0.003996642820031174, "grad_norm": 0.39886870980262756, "learning_rate": 2.4875000000000003e-06, "loss": 0.1559, "step": 200 }, { "epoch": 0.004196474961032732, "grad_norm": 0.35400450229644775, "learning_rate": 2.6125e-06, "loss": 0.1394, "step": 210 }, { "epoch": 0.004396307102034291, "grad_norm": 0.3248610198497772, "learning_rate": 2.7375e-06, "loss": 0.1458, "step": 220 }, { "epoch": 0.00459613924303585, "grad_norm": 0.3943769037723541, "learning_rate": 2.8625e-06, "loss": 0.1278, "step": 230 }, { "epoch": 0.004795971384037408, "grad_norm": 0.36580976843833923, "learning_rate": 2.9875e-06, "loss": 0.1385, "step": 240 }, { "epoch": 0.004995803525038967, "grad_norm": 0.39749613404273987, "learning_rate": 3.1125000000000003e-06, "loss": 0.1307, "step": 250 }, { "epoch": 0.005195635666040526, "grad_norm": 0.30624955892562866, "learning_rate": 3.2375000000000003e-06, "loss": 0.135, "step": 260 }, { "epoch": 0.005395467807042084, "grad_norm": 0.351127028465271, "learning_rate": 3.3625000000000004e-06, "loss": 0.1234, "step": 270 }, { "epoch": 0.005595299948043643, "grad_norm": 0.4367784559726715, "learning_rate": 3.4875000000000005e-06, "loss": 0.1392, "step": 280 }, { "epoch": 0.005795132089045202, "grad_norm": 0.32164326310157776, "learning_rate": 3.6124999999999997e-06, "loss": 0.1096, "step": 290 }, { "epoch": 0.0059949642300467605, "grad_norm": 0.35341447591781616, "learning_rate": 3.7375000000000002e-06, "loss": 0.122, "step": 300 }, { "epoch": 0.0061947963710483195, "grad_norm": 0.36957231163978577, "learning_rate": 3.8625e-06, "loss": 0.1145, "step": 310 }, { "epoch": 0.0063946285120498785, "grad_norm": 0.3842404782772064, "learning_rate": 3.9875e-06, "loss": 0.1196, "step": 320 }, { "epoch": 0.006594460653051437, "grad_norm": 0.3120656907558441, "learning_rate": 4.1125e-06, "loss": 0.1191, "step": 330 }, { "epoch": 0.006794292794052996, "grad_norm": 0.4664347767829895, "learning_rate": 4.2375000000000005e-06, "loss": 0.1051, "step": 340 }, { "epoch": 0.006994124935054554, "grad_norm": 0.394212007522583, "learning_rate": 4.3625e-06, "loss": 0.1162, "step": 350 }, { "epoch": 0.007193957076056113, "grad_norm": 0.36807650327682495, "learning_rate": 4.4875e-06, "loss": 0.1151, "step": 360 }, { "epoch": 0.007393789217057672, "grad_norm": 0.3324626684188843, "learning_rate": 4.6125e-06, "loss": 0.095, "step": 370 }, { "epoch": 0.00759362135805923, "grad_norm": 0.4082213342189789, "learning_rate": 4.7375e-06, "loss": 0.1101, "step": 380 }, { "epoch": 0.007793453499060789, "grad_norm": 0.3976486325263977, "learning_rate": 4.8625000000000005e-06, "loss": 0.1212, "step": 390 }, { "epoch": 0.007993285640062348, "grad_norm": 0.32062286138534546, "learning_rate": 4.9875e-06, "loss": 0.0963, "step": 400 }, { "epoch": 0.008193117781063906, "grad_norm": 0.325745552778244, "learning_rate": 5.1125e-06, "loss": 0.1008, "step": 410 }, { "epoch": 0.008392949922065464, "grad_norm": 0.334374338388443, "learning_rate": 5.2375e-06, "loss": 0.1146, "step": 420 }, { "epoch": 0.008592782063067024, "grad_norm": 0.37165367603302, "learning_rate": 5.3625e-06, "loss": 0.1061, "step": 430 }, { "epoch": 0.008792614204068582, "grad_norm": 0.3890312612056732, "learning_rate": 5.4875e-06, "loss": 0.1049, "step": 440 }, { "epoch": 0.00899244634507014, "grad_norm": 0.49855324625968933, "learning_rate": 5.6125e-06, "loss": 0.1016, "step": 450 }, { "epoch": 0.0091922784860717, "grad_norm": 0.3237970769405365, "learning_rate": 5.7375000000000005e-06, "loss": 0.0926, "step": 460 }, { "epoch": 0.009392110627073258, "grad_norm": 0.41110485792160034, "learning_rate": 5.8625e-06, "loss": 0.1032, "step": 470 }, { "epoch": 0.009591942768074816, "grad_norm": 0.3801605999469757, "learning_rate": 5.9875e-06, "loss": 0.0935, "step": 480 }, { "epoch": 0.009791774909076376, "grad_norm": 0.42626193165779114, "learning_rate": 6.1125e-06, "loss": 0.0996, "step": 490 }, { "epoch": 0.009991607050077934, "grad_norm": 0.4325594902038574, "learning_rate": 6.2375e-06, "loss": 0.0866, "step": 500 }, { "epoch": 0.010191439191079493, "grad_norm": 0.4070633053779602, "learning_rate": 6.3625000000000005e-06, "loss": 0.0932, "step": 510 }, { "epoch": 0.010391271332081052, "grad_norm": 0.47067275643348694, "learning_rate": 6.4875e-06, "loss": 0.0795, "step": 520 }, { "epoch": 0.01059110347308261, "grad_norm": 0.4359688460826874, "learning_rate": 6.612500000000001e-06, "loss": 0.1003, "step": 530 }, { "epoch": 0.010790935614084169, "grad_norm": 0.4188236594200134, "learning_rate": 6.737500000000001e-06, "loss": 0.1103, "step": 540 }, { "epoch": 0.010990767755085729, "grad_norm": 0.4163776934146881, "learning_rate": 6.862500000000001e-06, "loss": 0.0879, "step": 550 }, { "epoch": 0.011190599896087287, "grad_norm": 0.44475629925727844, "learning_rate": 6.987500000000001e-06, "loss": 0.0967, "step": 560 }, { "epoch": 0.011390432037088845, "grad_norm": 0.3383370339870453, "learning_rate": 7.1125e-06, "loss": 0.1077, "step": 570 }, { "epoch": 0.011590264178090405, "grad_norm": 0.4220973551273346, "learning_rate": 7.2375e-06, "loss": 0.1005, "step": 580 }, { "epoch": 0.011790096319091963, "grad_norm": 0.33477509021759033, "learning_rate": 7.3625e-06, "loss": 0.0824, "step": 590 }, { "epoch": 0.011989928460093521, "grad_norm": 0.4262847304344177, "learning_rate": 7.4875e-06, "loss": 0.0941, "step": 600 }, { "epoch": 0.01218976060109508, "grad_norm": 0.4283199906349182, "learning_rate": 7.6125e-06, "loss": 0.101, "step": 610 }, { "epoch": 0.012389592742096639, "grad_norm": 0.4418812692165375, "learning_rate": 7.7375e-06, "loss": 0.0946, "step": 620 }, { "epoch": 0.012589424883098197, "grad_norm": 0.32468700408935547, "learning_rate": 7.8625e-06, "loss": 0.0886, "step": 630 }, { "epoch": 0.012789257024099757, "grad_norm": 0.39358627796173096, "learning_rate": 7.987500000000001e-06, "loss": 0.0809, "step": 640 }, { "epoch": 0.012989089165101315, "grad_norm": 0.37120312452316284, "learning_rate": 8.1125e-06, "loss": 0.0865, "step": 650 }, { "epoch": 0.013188921306102873, "grad_norm": 0.3376367390155792, "learning_rate": 8.2375e-06, "loss": 0.0873, "step": 660 }, { "epoch": 0.013388753447104433, "grad_norm": 0.33446404337882996, "learning_rate": 8.362500000000002e-06, "loss": 0.0692, "step": 670 }, { "epoch": 0.013588585588105991, "grad_norm": 0.43137702345848083, "learning_rate": 8.487500000000001e-06, "loss": 0.0931, "step": 680 }, { "epoch": 0.01378841772910755, "grad_norm": 0.3259783089160919, "learning_rate": 8.6125e-06, "loss": 0.0748, "step": 690 }, { "epoch": 0.013988249870109108, "grad_norm": 0.37620967626571655, "learning_rate": 8.7375e-06, "loss": 0.0809, "step": 700 }, { "epoch": 0.014188082011110667, "grad_norm": 0.4211740493774414, "learning_rate": 8.8625e-06, "loss": 0.0883, "step": 710 }, { "epoch": 0.014387914152112226, "grad_norm": 0.4341007173061371, "learning_rate": 8.9875e-06, "loss": 0.0766, "step": 720 }, { "epoch": 0.014587746293113784, "grad_norm": 0.4108606278896332, "learning_rate": 9.1125e-06, "loss": 0.0916, "step": 730 }, { "epoch": 0.014787578434115344, "grad_norm": 0.3288455605506897, "learning_rate": 9.237500000000001e-06, "loss": 0.0796, "step": 740 }, { "epoch": 0.014987410575116902, "grad_norm": 0.3143172264099121, "learning_rate": 9.3625e-06, "loss": 0.0816, "step": 750 }, { "epoch": 0.01518724271611846, "grad_norm": 0.3082760274410248, "learning_rate": 9.4875e-06, "loss": 0.0805, "step": 760 }, { "epoch": 0.01538707485712002, "grad_norm": 0.3682020604610443, "learning_rate": 9.6125e-06, "loss": 0.0813, "step": 770 }, { "epoch": 0.015586906998121578, "grad_norm": 0.326013445854187, "learning_rate": 9.737500000000001e-06, "loss": 0.0846, "step": 780 }, { "epoch": 0.015786739139123138, "grad_norm": 0.3289579153060913, "learning_rate": 9.862500000000001e-06, "loss": 0.0702, "step": 790 }, { "epoch": 0.015986571280124696, "grad_norm": 0.3388408124446869, "learning_rate": 9.9875e-06, "loss": 0.0807, "step": 800 }, { "epoch": 0.016186403421126254, "grad_norm": 0.4479566514492035, "learning_rate": 1.0112500000000002e-05, "loss": 0.072, "step": 810 }, { "epoch": 0.016386235562127812, "grad_norm": 0.3914206027984619, "learning_rate": 1.02375e-05, "loss": 0.0967, "step": 820 }, { "epoch": 0.01658606770312937, "grad_norm": 0.30732983350753784, "learning_rate": 1.03625e-05, "loss": 0.0836, "step": 830 }, { "epoch": 0.01678589984413093, "grad_norm": 0.3702974021434784, "learning_rate": 1.04875e-05, "loss": 0.0713, "step": 840 }, { "epoch": 0.01698573198513249, "grad_norm": 0.46387678384780884, "learning_rate": 1.06125e-05, "loss": 0.0805, "step": 850 }, { "epoch": 0.017185564126134048, "grad_norm": 0.5607987642288208, "learning_rate": 1.07375e-05, "loss": 0.0898, "step": 860 }, { "epoch": 0.017385396267135606, "grad_norm": 0.41388753056526184, "learning_rate": 1.08625e-05, "loss": 0.0787, "step": 870 }, { "epoch": 0.017585228408137164, "grad_norm": 0.3108460307121277, "learning_rate": 1.0987500000000001e-05, "loss": 0.081, "step": 880 }, { "epoch": 0.017785060549138722, "grad_norm": 0.455668181180954, "learning_rate": 1.11125e-05, "loss": 0.0831, "step": 890 }, { "epoch": 0.01798489269014028, "grad_norm": 0.3034018576145172, "learning_rate": 1.12375e-05, "loss": 0.0784, "step": 900 }, { "epoch": 0.018184724831141842, "grad_norm": 0.3239409923553467, "learning_rate": 1.1362500000000002e-05, "loss": 0.0773, "step": 910 }, { "epoch": 0.0183845569721434, "grad_norm": 0.3783259987831116, "learning_rate": 1.1487500000000001e-05, "loss": 0.0832, "step": 920 }, { "epoch": 0.01858438911314496, "grad_norm": 0.3740631639957428, "learning_rate": 1.1612500000000001e-05, "loss": 0.0994, "step": 930 }, { "epoch": 0.018784221254146517, "grad_norm": 0.37053629755973816, "learning_rate": 1.17375e-05, "loss": 0.0683, "step": 940 }, { "epoch": 0.018984053395148075, "grad_norm": 0.5384717583656311, "learning_rate": 1.18625e-05, "loss": 0.0736, "step": 950 }, { "epoch": 0.019183885536149633, "grad_norm": 0.4366263151168823, "learning_rate": 1.19875e-05, "loss": 0.0849, "step": 960 }, { "epoch": 0.019383717677151194, "grad_norm": 0.3447225093841553, "learning_rate": 1.21125e-05, "loss": 0.0751, "step": 970 }, { "epoch": 0.019583549818152753, "grad_norm": 0.4051327705383301, "learning_rate": 1.2237500000000001e-05, "loss": 0.0786, "step": 980 }, { "epoch": 0.01978338195915431, "grad_norm": 0.3774309754371643, "learning_rate": 1.23625e-05, "loss": 0.0737, "step": 990 }, { "epoch": 0.01998321410015587, "grad_norm": 0.33270373940467834, "learning_rate": 1.24875e-05, "loss": 0.0718, "step": 1000 }, { "epoch": 0.020183046241157427, "grad_norm": 0.3553261458873749, "learning_rate": 1.26125e-05, "loss": 0.0883, "step": 1010 }, { "epoch": 0.020382878382158985, "grad_norm": 0.3042048215866089, "learning_rate": 1.27375e-05, "loss": 0.0713, "step": 1020 }, { "epoch": 0.020582710523160547, "grad_norm": 0.30412912368774414, "learning_rate": 1.28625e-05, "loss": 0.0745, "step": 1030 }, { "epoch": 0.020782542664162105, "grad_norm": 0.3008694648742676, "learning_rate": 1.2987499999999999e-05, "loss": 0.0706, "step": 1040 }, { "epoch": 0.020982374805163663, "grad_norm": 0.3785236179828644, "learning_rate": 1.31125e-05, "loss": 0.075, "step": 1050 }, { "epoch": 0.02118220694616522, "grad_norm": 0.2894957661628723, "learning_rate": 1.32375e-05, "loss": 0.0822, "step": 1060 }, { "epoch": 0.02138203908716678, "grad_norm": 0.3561248183250427, "learning_rate": 1.33625e-05, "loss": 0.0675, "step": 1070 }, { "epoch": 0.021581871228168337, "grad_norm": 0.29665708541870117, "learning_rate": 1.3487500000000001e-05, "loss": 0.0799, "step": 1080 }, { "epoch": 0.0217817033691699, "grad_norm": 0.3011758327484131, "learning_rate": 1.36125e-05, "loss": 0.0786, "step": 1090 }, { "epoch": 0.021981535510171457, "grad_norm": 0.5134522318840027, "learning_rate": 1.37375e-05, "loss": 0.073, "step": 1100 }, { "epoch": 0.022181367651173015, "grad_norm": 0.36947840452194214, "learning_rate": 1.38625e-05, "loss": 0.0689, "step": 1110 }, { "epoch": 0.022381199792174573, "grad_norm": 0.3480295240879059, "learning_rate": 1.3987500000000001e-05, "loss": 0.0716, "step": 1120 }, { "epoch": 0.02258103193317613, "grad_norm": 0.32793253660202026, "learning_rate": 1.41125e-05, "loss": 0.0748, "step": 1130 }, { "epoch": 0.02278086407417769, "grad_norm": 0.4468984007835388, "learning_rate": 1.42375e-05, "loss": 0.0649, "step": 1140 }, { "epoch": 0.022980696215179248, "grad_norm": 0.39618200063705444, "learning_rate": 1.43625e-05, "loss": 0.0825, "step": 1150 }, { "epoch": 0.02318052835618081, "grad_norm": 0.34320181608200073, "learning_rate": 1.4487500000000001e-05, "loss": 0.0679, "step": 1160 }, { "epoch": 0.023380360497182368, "grad_norm": 0.36643052101135254, "learning_rate": 1.4612500000000001e-05, "loss": 0.0774, "step": 1170 }, { "epoch": 0.023580192638183926, "grad_norm": 0.3313843011856079, "learning_rate": 1.47375e-05, "loss": 0.0755, "step": 1180 }, { "epoch": 0.023780024779185484, "grad_norm": 0.33172607421875, "learning_rate": 1.4862500000000002e-05, "loss": 0.0771, "step": 1190 }, { "epoch": 0.023979856920187042, "grad_norm": 0.36656978726387024, "learning_rate": 1.4987500000000002e-05, "loss": 0.0777, "step": 1200 }, { "epoch": 0.0241796890611886, "grad_norm": 0.2312343716621399, "learning_rate": 1.5112500000000001e-05, "loss": 0.0867, "step": 1210 }, { "epoch": 0.02437952120219016, "grad_norm": 0.3517617881298065, "learning_rate": 1.5237500000000001e-05, "loss": 0.0823, "step": 1220 }, { "epoch": 0.02457935334319172, "grad_norm": 0.3248494565486908, "learning_rate": 1.53625e-05, "loss": 0.0675, "step": 1230 }, { "epoch": 0.024779185484193278, "grad_norm": 0.5002638101577759, "learning_rate": 1.54875e-05, "loss": 0.0739, "step": 1240 }, { "epoch": 0.024979017625194836, "grad_norm": 0.29400280117988586, "learning_rate": 1.5612500000000003e-05, "loss": 0.065, "step": 1250 }, { "epoch": 0.025178849766196394, "grad_norm": 0.3151928186416626, "learning_rate": 1.57375e-05, "loss": 0.0662, "step": 1260 }, { "epoch": 0.025378681907197952, "grad_norm": 0.3570931851863861, "learning_rate": 1.58625e-05, "loss": 0.0625, "step": 1270 }, { "epoch": 0.025578514048199514, "grad_norm": 0.33401989936828613, "learning_rate": 1.59875e-05, "loss": 0.0753, "step": 1280 }, { "epoch": 0.025778346189201072, "grad_norm": 0.3104116916656494, "learning_rate": 1.61125e-05, "loss": 0.0713, "step": 1290 }, { "epoch": 0.02597817833020263, "grad_norm": 0.31080150604248047, "learning_rate": 1.6237499999999998e-05, "loss": 0.0746, "step": 1300 }, { "epoch": 0.02617801047120419, "grad_norm": 0.2771340012550354, "learning_rate": 1.63625e-05, "loss": 0.0672, "step": 1310 }, { "epoch": 0.026377842612205746, "grad_norm": 0.33741044998168945, "learning_rate": 1.64875e-05, "loss": 0.073, "step": 1320 }, { "epoch": 0.026577674753207305, "grad_norm": 0.3514914810657501, "learning_rate": 1.66125e-05, "loss": 0.0768, "step": 1330 }, { "epoch": 0.026777506894208866, "grad_norm": 0.36171436309814453, "learning_rate": 1.67375e-05, "loss": 0.0919, "step": 1340 }, { "epoch": 0.026977339035210424, "grad_norm": 0.32689833641052246, "learning_rate": 1.68625e-05, "loss": 0.0678, "step": 1350 }, { "epoch": 0.027177171176211982, "grad_norm": 0.39326220750808716, "learning_rate": 1.69875e-05, "loss": 0.0814, "step": 1360 }, { "epoch": 0.02737700331721354, "grad_norm": 0.3633366525173187, "learning_rate": 1.71125e-05, "loss": 0.0692, "step": 1370 }, { "epoch": 0.0275768354582151, "grad_norm": 0.46718043088912964, "learning_rate": 1.7237500000000002e-05, "loss": 0.0729, "step": 1380 }, { "epoch": 0.027776667599216657, "grad_norm": 0.3228197991847992, "learning_rate": 1.7362500000000002e-05, "loss": 0.0733, "step": 1390 }, { "epoch": 0.027976499740218215, "grad_norm": 0.349138081073761, "learning_rate": 1.74875e-05, "loss": 0.0711, "step": 1400 }, { "epoch": 0.028176331881219777, "grad_norm": 0.35699206590652466, "learning_rate": 1.76125e-05, "loss": 0.0856, "step": 1410 }, { "epoch": 0.028376164022221335, "grad_norm": 0.3093022108078003, "learning_rate": 1.77375e-05, "loss": 0.0758, "step": 1420 }, { "epoch": 0.028575996163222893, "grad_norm": 0.37190860509872437, "learning_rate": 1.78625e-05, "loss": 0.0768, "step": 1430 }, { "epoch": 0.02877582830422445, "grad_norm": 0.4406457543373108, "learning_rate": 1.79875e-05, "loss": 0.0645, "step": 1440 }, { "epoch": 0.02897566044522601, "grad_norm": 0.3963196575641632, "learning_rate": 1.8112500000000003e-05, "loss": 0.0754, "step": 1450 }, { "epoch": 0.029175492586227567, "grad_norm": 0.3091088831424713, "learning_rate": 1.8237500000000003e-05, "loss": 0.069, "step": 1460 }, { "epoch": 0.02937532472722913, "grad_norm": 0.2834456264972687, "learning_rate": 1.8362500000000002e-05, "loss": 0.0584, "step": 1470 }, { "epoch": 0.029575156868230687, "grad_norm": 0.38622522354125977, "learning_rate": 1.8487500000000002e-05, "loss": 0.0642, "step": 1480 }, { "epoch": 0.029774989009232245, "grad_norm": 0.41308391094207764, "learning_rate": 1.8612500000000002e-05, "loss": 0.0664, "step": 1490 }, { "epoch": 0.029974821150233803, "grad_norm": 0.3075287938117981, "learning_rate": 1.87375e-05, "loss": 0.0779, "step": 1500 }, { "epoch": 0.03017465329123536, "grad_norm": 0.32369473576545715, "learning_rate": 1.88625e-05, "loss": 0.0597, "step": 1510 }, { "epoch": 0.03037448543223692, "grad_norm": 0.2913534939289093, "learning_rate": 1.89875e-05, "loss": 0.0657, "step": 1520 }, { "epoch": 0.03057431757323848, "grad_norm": 0.30214497447013855, "learning_rate": 1.91125e-05, "loss": 0.0713, "step": 1530 }, { "epoch": 0.03077414971424004, "grad_norm": 0.33609527349472046, "learning_rate": 1.92375e-05, "loss": 0.0625, "step": 1540 }, { "epoch": 0.030973981855241597, "grad_norm": 0.37011241912841797, "learning_rate": 1.93625e-05, "loss": 0.0799, "step": 1550 }, { "epoch": 0.031173813996243156, "grad_norm": 0.3026440739631653, "learning_rate": 1.94875e-05, "loss": 0.0571, "step": 1560 }, { "epoch": 0.031373646137244714, "grad_norm": 0.27416086196899414, "learning_rate": 1.96125e-05, "loss": 0.0646, "step": 1570 }, { "epoch": 0.031573478278246275, "grad_norm": 0.41401150822639465, "learning_rate": 1.9737500000000002e-05, "loss": 0.0693, "step": 1580 }, { "epoch": 0.03177331041924783, "grad_norm": 0.36536312103271484, "learning_rate": 1.98625e-05, "loss": 0.0693, "step": 1590 }, { "epoch": 0.03197314256024939, "grad_norm": 0.2900424301624298, "learning_rate": 1.99875e-05, "loss": 0.0691, "step": 1600 }, { "epoch": 0.032172974701250946, "grad_norm": 0.327888548374176, "learning_rate": 2.01125e-05, "loss": 0.0743, "step": 1610 }, { "epoch": 0.03237280684225251, "grad_norm": 0.27838847041130066, "learning_rate": 2.02375e-05, "loss": 0.069, "step": 1620 }, { "epoch": 0.03257263898325407, "grad_norm": 0.36008426547050476, "learning_rate": 2.03625e-05, "loss": 0.0755, "step": 1630 }, { "epoch": 0.032772471124255624, "grad_norm": 0.3451668322086334, "learning_rate": 2.04875e-05, "loss": 0.0656, "step": 1640 }, { "epoch": 0.032972303265257186, "grad_norm": 0.28650107979774475, "learning_rate": 2.06125e-05, "loss": 0.0647, "step": 1650 }, { "epoch": 0.03317213540625874, "grad_norm": 0.28486964106559753, "learning_rate": 2.0737500000000003e-05, "loss": 0.0762, "step": 1660 }, { "epoch": 0.0333719675472603, "grad_norm": 0.3788469135761261, "learning_rate": 2.0862500000000002e-05, "loss": 0.0652, "step": 1670 }, { "epoch": 0.03357179968826186, "grad_norm": 0.31422480940818787, "learning_rate": 2.0987500000000002e-05, "loss": 0.0619, "step": 1680 }, { "epoch": 0.03377163182926342, "grad_norm": 0.29629865288734436, "learning_rate": 2.11125e-05, "loss": 0.081, "step": 1690 }, { "epoch": 0.03397146397026498, "grad_norm": 0.2994856834411621, "learning_rate": 2.12375e-05, "loss": 0.0668, "step": 1700 }, { "epoch": 0.034171296111266534, "grad_norm": 0.2729422152042389, "learning_rate": 2.13625e-05, "loss": 0.0646, "step": 1710 }, { "epoch": 0.034371128252268096, "grad_norm": 0.39032605290412903, "learning_rate": 2.14875e-05, "loss": 0.0728, "step": 1720 }, { "epoch": 0.03457096039326965, "grad_norm": 0.34050384163856506, "learning_rate": 2.1612500000000004e-05, "loss": 0.0598, "step": 1730 }, { "epoch": 0.03477079253427121, "grad_norm": 0.2876433730125427, "learning_rate": 2.1737500000000003e-05, "loss": 0.0745, "step": 1740 }, { "epoch": 0.034970624675272774, "grad_norm": 0.28272780776023865, "learning_rate": 2.1862500000000003e-05, "loss": 0.0618, "step": 1750 }, { "epoch": 0.03517045681627433, "grad_norm": 0.2799200117588043, "learning_rate": 2.19875e-05, "loss": 0.0665, "step": 1760 }, { "epoch": 0.03537028895727589, "grad_norm": 0.3617914319038391, "learning_rate": 2.21125e-05, "loss": 0.0654, "step": 1770 }, { "epoch": 0.035570121098277445, "grad_norm": 0.306878000497818, "learning_rate": 2.22375e-05, "loss": 0.057, "step": 1780 }, { "epoch": 0.035769953239279006, "grad_norm": 0.24603183567523956, "learning_rate": 2.23625e-05, "loss": 0.0737, "step": 1790 }, { "epoch": 0.03596978538028056, "grad_norm": 0.29952874779701233, "learning_rate": 2.24875e-05, "loss": 0.0703, "step": 1800 }, { "epoch": 0.03616961752128212, "grad_norm": 0.30930930376052856, "learning_rate": 2.26125e-05, "loss": 0.0634, "step": 1810 }, { "epoch": 0.036369449662283684, "grad_norm": 0.2964021563529968, "learning_rate": 2.27375e-05, "loss": 0.0537, "step": 1820 }, { "epoch": 0.03656928180328524, "grad_norm": 0.291299045085907, "learning_rate": 2.28625e-05, "loss": 0.0684, "step": 1830 }, { "epoch": 0.0367691139442868, "grad_norm": 0.3443099558353424, "learning_rate": 2.29875e-05, "loss": 0.0717, "step": 1840 }, { "epoch": 0.036968946085288355, "grad_norm": 0.5796244740486145, "learning_rate": 2.31125e-05, "loss": 0.0626, "step": 1850 }, { "epoch": 0.03716877822628992, "grad_norm": 0.3886847496032715, "learning_rate": 2.3237500000000002e-05, "loss": 0.0662, "step": 1860 }, { "epoch": 0.03736861036729148, "grad_norm": 0.3061986565589905, "learning_rate": 2.3362500000000002e-05, "loss": 0.0598, "step": 1870 }, { "epoch": 0.03756844250829303, "grad_norm": 0.2492590844631195, "learning_rate": 2.3487500000000002e-05, "loss": 0.0615, "step": 1880 }, { "epoch": 0.037768274649294595, "grad_norm": 0.25430625677108765, "learning_rate": 2.36125e-05, "loss": 0.0728, "step": 1890 }, { "epoch": 0.03796810679029615, "grad_norm": 0.25509318709373474, "learning_rate": 2.37375e-05, "loss": 0.0709, "step": 1900 }, { "epoch": 0.03816793893129771, "grad_norm": 0.35073065757751465, "learning_rate": 2.38625e-05, "loss": 0.0648, "step": 1910 }, { "epoch": 0.038367771072299266, "grad_norm": 0.4106261134147644, "learning_rate": 2.39875e-05, "loss": 0.0544, "step": 1920 }, { "epoch": 0.03856760321330083, "grad_norm": 0.331495463848114, "learning_rate": 2.4112500000000003e-05, "loss": 0.0694, "step": 1930 }, { "epoch": 0.03876743535430239, "grad_norm": 0.2975996136665344, "learning_rate": 2.4237500000000003e-05, "loss": 0.0568, "step": 1940 }, { "epoch": 0.038967267495303944, "grad_norm": 0.27191823720932007, "learning_rate": 2.4362500000000003e-05, "loss": 0.0702, "step": 1950 }, { "epoch": 0.039167099636305505, "grad_norm": 0.46075889468193054, "learning_rate": 2.4487500000000002e-05, "loss": 0.0673, "step": 1960 }, { "epoch": 0.03936693177730706, "grad_norm": 0.2594573199748993, "learning_rate": 2.4612500000000002e-05, "loss": 0.0663, "step": 1970 }, { "epoch": 0.03956676391830862, "grad_norm": 0.3970872759819031, "learning_rate": 2.47375e-05, "loss": 0.068, "step": 1980 }, { "epoch": 0.039766596059310176, "grad_norm": 0.327825129032135, "learning_rate": 2.48625e-05, "loss": 0.0598, "step": 1990 }, { "epoch": 0.03996642820031174, "grad_norm": 0.44912832975387573, "learning_rate": 2.49875e-05, "loss": 0.072, "step": 2000 }, { "epoch": 0.0401662603413133, "grad_norm": 0.3663671016693115, "learning_rate": 2.51125e-05, "loss": 0.0628, "step": 2010 }, { "epoch": 0.040366092482314854, "grad_norm": 0.24780720472335815, "learning_rate": 2.5237500000000004e-05, "loss": 0.0567, "step": 2020 }, { "epoch": 0.040565924623316416, "grad_norm": 0.3126686215400696, "learning_rate": 2.53625e-05, "loss": 0.0649, "step": 2030 }, { "epoch": 0.04076575676431797, "grad_norm": 0.29543641209602356, "learning_rate": 2.5487500000000003e-05, "loss": 0.06, "step": 2040 }, { "epoch": 0.04096558890531953, "grad_norm": 0.3644212484359741, "learning_rate": 2.56125e-05, "loss": 0.0613, "step": 2050 }, { "epoch": 0.04116542104632109, "grad_norm": 0.28476396203041077, "learning_rate": 2.5737500000000002e-05, "loss": 0.0655, "step": 2060 }, { "epoch": 0.04136525318732265, "grad_norm": 0.2922973036766052, "learning_rate": 2.5862500000000002e-05, "loss": 0.0671, "step": 2070 }, { "epoch": 0.04156508532832421, "grad_norm": 0.307646244764328, "learning_rate": 2.5987500000000005e-05, "loss": 0.0671, "step": 2080 }, { "epoch": 0.041764917469325764, "grad_norm": 0.3630848228931427, "learning_rate": 2.61125e-05, "loss": 0.0591, "step": 2090 }, { "epoch": 0.041964749610327326, "grad_norm": 0.24032413959503174, "learning_rate": 2.6237500000000004e-05, "loss": 0.0611, "step": 2100 }, { "epoch": 0.04216458175132888, "grad_norm": 0.21118630468845367, "learning_rate": 2.63625e-05, "loss": 0.0665, "step": 2110 }, { "epoch": 0.04236441389233044, "grad_norm": 0.290635347366333, "learning_rate": 2.6487500000000003e-05, "loss": 0.0791, "step": 2120 }, { "epoch": 0.042564246033332004, "grad_norm": 0.30253201723098755, "learning_rate": 2.66125e-05, "loss": 0.0711, "step": 2130 }, { "epoch": 0.04276407817433356, "grad_norm": 0.3474222421646118, "learning_rate": 2.67375e-05, "loss": 0.0655, "step": 2140 }, { "epoch": 0.04296391031533512, "grad_norm": 0.33708274364471436, "learning_rate": 2.6862500000000002e-05, "loss": 0.0603, "step": 2150 }, { "epoch": 0.043163742456336675, "grad_norm": 0.3444520831108093, "learning_rate": 2.69875e-05, "loss": 0.0739, "step": 2160 }, { "epoch": 0.043363574597338236, "grad_norm": 0.25322332978248596, "learning_rate": 2.7112500000000002e-05, "loss": 0.0529, "step": 2170 }, { "epoch": 0.0435634067383398, "grad_norm": 0.2187170684337616, "learning_rate": 2.7237499999999998e-05, "loss": 0.0513, "step": 2180 }, { "epoch": 0.04376323887934135, "grad_norm": 0.2596800625324249, "learning_rate": 2.73625e-05, "loss": 0.0659, "step": 2190 }, { "epoch": 0.043963071020342914, "grad_norm": 0.24219611287117004, "learning_rate": 2.74875e-05, "loss": 0.0608, "step": 2200 }, { "epoch": 0.04416290316134447, "grad_norm": 0.27955949306488037, "learning_rate": 2.7612500000000004e-05, "loss": 0.0707, "step": 2210 }, { "epoch": 0.04436273530234603, "grad_norm": 0.3381091058254242, "learning_rate": 2.77375e-05, "loss": 0.0662, "step": 2220 }, { "epoch": 0.044562567443347585, "grad_norm": 0.3584311008453369, "learning_rate": 2.7862500000000003e-05, "loss": 0.0678, "step": 2230 }, { "epoch": 0.04476239958434915, "grad_norm": 0.26175326108932495, "learning_rate": 2.79875e-05, "loss": 0.0625, "step": 2240 }, { "epoch": 0.04496223172535071, "grad_norm": 0.253462016582489, "learning_rate": 2.8112500000000002e-05, "loss": 0.0656, "step": 2250 }, { "epoch": 0.04516206386635226, "grad_norm": 0.2404707819223404, "learning_rate": 2.82375e-05, "loss": 0.0743, "step": 2260 }, { "epoch": 0.045361896007353825, "grad_norm": 0.24193012714385986, "learning_rate": 2.83625e-05, "loss": 0.0673, "step": 2270 }, { "epoch": 0.04556172814835538, "grad_norm": 0.24507258832454681, "learning_rate": 2.84875e-05, "loss": 0.0646, "step": 2280 }, { "epoch": 0.04576156028935694, "grad_norm": 0.2718161344528198, "learning_rate": 2.8612500000000004e-05, "loss": 0.068, "step": 2290 }, { "epoch": 0.045961392430358496, "grad_norm": 0.3183963894844055, "learning_rate": 2.87375e-05, "loss": 0.067, "step": 2300 }, { "epoch": 0.04616122457136006, "grad_norm": 0.36428290605545044, "learning_rate": 2.8862500000000004e-05, "loss": 0.0573, "step": 2310 }, { "epoch": 0.04636105671236162, "grad_norm": 0.27712464332580566, "learning_rate": 2.89875e-05, "loss": 0.0659, "step": 2320 }, { "epoch": 0.04656088885336317, "grad_norm": 0.3745790421962738, "learning_rate": 2.9112500000000003e-05, "loss": 0.0636, "step": 2330 }, { "epoch": 0.046760720994364735, "grad_norm": 0.23631535470485687, "learning_rate": 2.9237500000000003e-05, "loss": 0.0648, "step": 2340 }, { "epoch": 0.04696055313536629, "grad_norm": 0.3086974024772644, "learning_rate": 2.9362500000000002e-05, "loss": 0.0653, "step": 2350 }, { "epoch": 0.04716038527636785, "grad_norm": 0.21463406085968018, "learning_rate": 2.9487500000000002e-05, "loss": 0.0609, "step": 2360 }, { "epoch": 0.04736021741736941, "grad_norm": 0.27037492394447327, "learning_rate": 2.9612500000000005e-05, "loss": 0.0675, "step": 2370 }, { "epoch": 0.04756004955837097, "grad_norm": 0.277713805437088, "learning_rate": 2.97375e-05, "loss": 0.0615, "step": 2380 }, { "epoch": 0.04775988169937253, "grad_norm": 0.365775465965271, "learning_rate": 2.9862499999999997e-05, "loss": 0.0741, "step": 2390 }, { "epoch": 0.047959713840374084, "grad_norm": 0.29813140630722046, "learning_rate": 2.99875e-05, "loss": 0.0591, "step": 2400 }, { "epoch": 0.048159545981375645, "grad_norm": 0.36092716455459595, "learning_rate": 3.01125e-05, "loss": 0.0686, "step": 2410 }, { "epoch": 0.0483593781223772, "grad_norm": 0.2775658667087555, "learning_rate": 3.0237500000000003e-05, "loss": 0.0647, "step": 2420 }, { "epoch": 0.04855921026337876, "grad_norm": 0.2691895067691803, "learning_rate": 3.03625e-05, "loss": 0.0604, "step": 2430 }, { "epoch": 0.04875904240438032, "grad_norm": 0.3101370334625244, "learning_rate": 3.0487500000000002e-05, "loss": 0.0622, "step": 2440 }, { "epoch": 0.04895887454538188, "grad_norm": 0.2466396689414978, "learning_rate": 3.06125e-05, "loss": 0.0655, "step": 2450 }, { "epoch": 0.04915870668638344, "grad_norm": 0.275082528591156, "learning_rate": 3.07375e-05, "loss": 0.0661, "step": 2460 }, { "epoch": 0.049358538827384994, "grad_norm": 0.3069266080856323, "learning_rate": 3.08625e-05, "loss": 0.0599, "step": 2470 }, { "epoch": 0.049558370968386556, "grad_norm": 0.2003774344921112, "learning_rate": 3.09875e-05, "loss": 0.0645, "step": 2480 }, { "epoch": 0.04975820310938811, "grad_norm": 0.35406172275543213, "learning_rate": 3.11125e-05, "loss": 0.0663, "step": 2490 }, { "epoch": 0.04995803525038967, "grad_norm": 0.24946372210979462, "learning_rate": 3.12375e-05, "loss": 0.0655, "step": 2500 }, { "epoch": 0.050157867391391234, "grad_norm": 0.2619459331035614, "learning_rate": 3.13625e-05, "loss": 0.0673, "step": 2510 }, { "epoch": 0.05035769953239279, "grad_norm": 0.3383641541004181, "learning_rate": 3.1487500000000006e-05, "loss": 0.0549, "step": 2520 }, { "epoch": 0.05055753167339435, "grad_norm": 0.2861096262931824, "learning_rate": 3.16125e-05, "loss": 0.0605, "step": 2530 }, { "epoch": 0.050757363814395905, "grad_norm": 0.28838253021240234, "learning_rate": 3.1737500000000006e-05, "loss": 0.0797, "step": 2540 }, { "epoch": 0.050957195955397466, "grad_norm": 0.2230621576309204, "learning_rate": 3.18625e-05, "loss": 0.0575, "step": 2550 }, { "epoch": 0.05115702809639903, "grad_norm": 0.25791510939598083, "learning_rate": 3.1987500000000005e-05, "loss": 0.0724, "step": 2560 }, { "epoch": 0.05135686023740058, "grad_norm": 0.33693811297416687, "learning_rate": 3.21125e-05, "loss": 0.0667, "step": 2570 }, { "epoch": 0.051556692378402144, "grad_norm": 0.341573566198349, "learning_rate": 3.2237500000000004e-05, "loss": 0.0699, "step": 2580 }, { "epoch": 0.0517565245194037, "grad_norm": 0.33872491121292114, "learning_rate": 3.23625e-05, "loss": 0.0611, "step": 2590 }, { "epoch": 0.05195635666040526, "grad_norm": 0.22042801976203918, "learning_rate": 3.2487500000000004e-05, "loss": 0.0655, "step": 2600 }, { "epoch": 0.052156188801406815, "grad_norm": 0.28331872820854187, "learning_rate": 3.26125e-05, "loss": 0.0651, "step": 2610 }, { "epoch": 0.05235602094240838, "grad_norm": 0.20906616747379303, "learning_rate": 3.27375e-05, "loss": 0.057, "step": 2620 }, { "epoch": 0.05255585308340994, "grad_norm": 0.29346731305122375, "learning_rate": 3.28625e-05, "loss": 0.0681, "step": 2630 }, { "epoch": 0.05275568522441149, "grad_norm": 0.2754264175891876, "learning_rate": 3.29875e-05, "loss": 0.0681, "step": 2640 }, { "epoch": 0.052955517365413055, "grad_norm": 0.20569495856761932, "learning_rate": 3.31125e-05, "loss": 0.0599, "step": 2650 }, { "epoch": 0.05315534950641461, "grad_norm": 0.25618839263916016, "learning_rate": 3.32375e-05, "loss": 0.0581, "step": 2660 }, { "epoch": 0.05335518164741617, "grad_norm": 0.24289736151695251, "learning_rate": 3.3362500000000005e-05, "loss": 0.069, "step": 2670 }, { "epoch": 0.05355501378841773, "grad_norm": 0.2187562882900238, "learning_rate": 3.34875e-05, "loss": 0.0581, "step": 2680 }, { "epoch": 0.05375484592941929, "grad_norm": 0.2613722085952759, "learning_rate": 3.3612500000000004e-05, "loss": 0.0536, "step": 2690 }, { "epoch": 0.05395467807042085, "grad_norm": 0.28154686093330383, "learning_rate": 3.37375e-05, "loss": 0.0611, "step": 2700 }, { "epoch": 0.0541545102114224, "grad_norm": 0.19883008301258087, "learning_rate": 3.38625e-05, "loss": 0.0568, "step": 2710 }, { "epoch": 0.054354342352423965, "grad_norm": 0.3000001311302185, "learning_rate": 3.39875e-05, "loss": 0.0664, "step": 2720 }, { "epoch": 0.05455417449342552, "grad_norm": 0.33956098556518555, "learning_rate": 3.41125e-05, "loss": 0.0514, "step": 2730 }, { "epoch": 0.05475400663442708, "grad_norm": 0.30166900157928467, "learning_rate": 3.42375e-05, "loss": 0.0588, "step": 2740 }, { "epoch": 0.05495383877542864, "grad_norm": 0.27743834257125854, "learning_rate": 3.43625e-05, "loss": 0.0633, "step": 2750 }, { "epoch": 0.0551536709164302, "grad_norm": 0.2563210427761078, "learning_rate": 3.44875e-05, "loss": 0.0631, "step": 2760 }, { "epoch": 0.05535350305743176, "grad_norm": 0.28725382685661316, "learning_rate": 3.46125e-05, "loss": 0.0552, "step": 2770 }, { "epoch": 0.055553335198433314, "grad_norm": 0.32145577669143677, "learning_rate": 3.47375e-05, "loss": 0.0563, "step": 2780 }, { "epoch": 0.055753167339434875, "grad_norm": 0.27794215083122253, "learning_rate": 3.48625e-05, "loss": 0.0682, "step": 2790 }, { "epoch": 0.05595299948043643, "grad_norm": 0.29540199041366577, "learning_rate": 3.4987500000000003e-05, "loss": 0.0617, "step": 2800 }, { "epoch": 0.05615283162143799, "grad_norm": 0.3548769950866699, "learning_rate": 3.5112500000000006e-05, "loss": 0.0593, "step": 2810 }, { "epoch": 0.05635266376243955, "grad_norm": 0.2125396430492401, "learning_rate": 3.52375e-05, "loss": 0.0548, "step": 2820 }, { "epoch": 0.05655249590344111, "grad_norm": 0.19852052628993988, "learning_rate": 3.5362500000000006e-05, "loss": 0.0672, "step": 2830 }, { "epoch": 0.05675232804444267, "grad_norm": 0.4029194414615631, "learning_rate": 3.54875e-05, "loss": 0.0634, "step": 2840 }, { "epoch": 0.056952160185444224, "grad_norm": 0.22278259694576263, "learning_rate": 3.5612500000000005e-05, "loss": 0.0604, "step": 2850 }, { "epoch": 0.057151992326445786, "grad_norm": 0.22452911734580994, "learning_rate": 3.57375e-05, "loss": 0.0627, "step": 2860 }, { "epoch": 0.05735182446744735, "grad_norm": 0.34067580103874207, "learning_rate": 3.5862500000000004e-05, "loss": 0.0685, "step": 2870 }, { "epoch": 0.0575516566084489, "grad_norm": 0.2717120349407196, "learning_rate": 3.59875e-05, "loss": 0.0546, "step": 2880 }, { "epoch": 0.057751488749450464, "grad_norm": 0.23335224390029907, "learning_rate": 3.61125e-05, "loss": 0.0685, "step": 2890 }, { "epoch": 0.05795132089045202, "grad_norm": 0.2513030171394348, "learning_rate": 3.62375e-05, "loss": 0.0614, "step": 2900 }, { "epoch": 0.05815115303145358, "grad_norm": 0.26043540239334106, "learning_rate": 3.6362499999999996e-05, "loss": 0.0589, "step": 2910 }, { "epoch": 0.058350985172455135, "grad_norm": 0.3057592511177063, "learning_rate": 3.64875e-05, "loss": 0.0639, "step": 2920 }, { "epoch": 0.058550817313456696, "grad_norm": 0.22490844130516052, "learning_rate": 3.66125e-05, "loss": 0.0599, "step": 2930 }, { "epoch": 0.05875064945445826, "grad_norm": 0.2781440317630768, "learning_rate": 3.6737500000000005e-05, "loss": 0.0675, "step": 2940 }, { "epoch": 0.05895048159545981, "grad_norm": 0.24750661849975586, "learning_rate": 3.68625e-05, "loss": 0.0684, "step": 2950 }, { "epoch": 0.059150313736461374, "grad_norm": 0.31198880076408386, "learning_rate": 3.6987500000000005e-05, "loss": 0.0683, "step": 2960 }, { "epoch": 0.05935014587746293, "grad_norm": 0.3363429307937622, "learning_rate": 3.71125e-05, "loss": 0.0625, "step": 2970 }, { "epoch": 0.05954997801846449, "grad_norm": 0.3044412136077881, "learning_rate": 3.7237500000000004e-05, "loss": 0.0576, "step": 2980 }, { "epoch": 0.05974981015946605, "grad_norm": 0.28388550877571106, "learning_rate": 3.73625e-05, "loss": 0.0653, "step": 2990 }, { "epoch": 0.05994964230046761, "grad_norm": 0.25386032462120056, "learning_rate": 3.74875e-05, "loss": 0.0782, "step": 3000 }, { "epoch": 0.06014947444146917, "grad_norm": 0.2545050382614136, "learning_rate": 3.76125e-05, "loss": 0.062, "step": 3010 }, { "epoch": 0.06034930658247072, "grad_norm": 0.3613395392894745, "learning_rate": 3.77375e-05, "loss": 0.065, "step": 3020 }, { "epoch": 0.060549138723472284, "grad_norm": 0.23558147251605988, "learning_rate": 3.78625e-05, "loss": 0.061, "step": 3030 }, { "epoch": 0.06074897086447384, "grad_norm": 0.3234497606754303, "learning_rate": 3.79875e-05, "loss": 0.0695, "step": 3040 }, { "epoch": 0.0609488030054754, "grad_norm": 0.25519147515296936, "learning_rate": 3.81125e-05, "loss": 0.0693, "step": 3050 }, { "epoch": 0.06114863514647696, "grad_norm": 0.22651739418506622, "learning_rate": 3.82375e-05, "loss": 0.0531, "step": 3060 }, { "epoch": 0.06134846728747852, "grad_norm": 0.28325286507606506, "learning_rate": 3.8362500000000004e-05, "loss": 0.0582, "step": 3070 }, { "epoch": 0.06154829942848008, "grad_norm": 0.33756744861602783, "learning_rate": 3.848750000000001e-05, "loss": 0.0811, "step": 3080 }, { "epoch": 0.06174813156948163, "grad_norm": 0.28434687852859497, "learning_rate": 3.8612500000000003e-05, "loss": 0.0627, "step": 3090 }, { "epoch": 0.061947963710483195, "grad_norm": 0.25095289945602417, "learning_rate": 3.8737500000000006e-05, "loss": 0.0631, "step": 3100 }, { "epoch": 0.06214779585148475, "grad_norm": 0.3749754726886749, "learning_rate": 3.88625e-05, "loss": 0.0616, "step": 3110 }, { "epoch": 0.06234762799248631, "grad_norm": 0.3309432864189148, "learning_rate": 3.8987500000000006e-05, "loss": 0.0631, "step": 3120 }, { "epoch": 0.06254746013348787, "grad_norm": 0.2778887450695038, "learning_rate": 3.91125e-05, "loss": 0.0593, "step": 3130 }, { "epoch": 0.06274729227448943, "grad_norm": 0.3019864559173584, "learning_rate": 3.92375e-05, "loss": 0.0595, "step": 3140 }, { "epoch": 0.06294712441549098, "grad_norm": 0.2471337914466858, "learning_rate": 3.93625e-05, "loss": 0.0604, "step": 3150 }, { "epoch": 0.06314695655649255, "grad_norm": 0.2988860011100769, "learning_rate": 3.94875e-05, "loss": 0.058, "step": 3160 }, { "epoch": 0.0633467886974941, "grad_norm": 0.22462980449199677, "learning_rate": 3.96125e-05, "loss": 0.0658, "step": 3170 }, { "epoch": 0.06354662083849566, "grad_norm": 0.2549641728401184, "learning_rate": 3.97375e-05, "loss": 0.0615, "step": 3180 }, { "epoch": 0.06374645297949723, "grad_norm": 0.26593154668807983, "learning_rate": 3.98625e-05, "loss": 0.0562, "step": 3190 }, { "epoch": 0.06394628512049878, "grad_norm": 0.2694603502750397, "learning_rate": 3.99875e-05, "loss": 0.0702, "step": 3200 }, { "epoch": 0.06414611726150034, "grad_norm": 0.3311172425746918, "learning_rate": 4.0112500000000006e-05, "loss": 0.0567, "step": 3210 }, { "epoch": 0.06434594940250189, "grad_norm": 0.31422626972198486, "learning_rate": 4.02375e-05, "loss": 0.0574, "step": 3220 }, { "epoch": 0.06454578154350346, "grad_norm": 0.1941489726305008, "learning_rate": 4.0362500000000005e-05, "loss": 0.0554, "step": 3230 }, { "epoch": 0.06474561368450502, "grad_norm": 0.2566578686237335, "learning_rate": 4.04875e-05, "loss": 0.0567, "step": 3240 }, { "epoch": 0.06494544582550657, "grad_norm": 0.2265915721654892, "learning_rate": 4.0612500000000005e-05, "loss": 0.0623, "step": 3250 }, { "epoch": 0.06514527796650814, "grad_norm": 0.29557234048843384, "learning_rate": 4.07375e-05, "loss": 0.0545, "step": 3260 }, { "epoch": 0.0653451101075097, "grad_norm": 0.23138785362243652, "learning_rate": 4.0862500000000004e-05, "loss": 0.0648, "step": 3270 }, { "epoch": 0.06554494224851125, "grad_norm": 0.22793643176555634, "learning_rate": 4.09875e-05, "loss": 0.0615, "step": 3280 }, { "epoch": 0.0657447743895128, "grad_norm": 0.258842408657074, "learning_rate": 4.11125e-05, "loss": 0.0599, "step": 3290 }, { "epoch": 0.06594460653051437, "grad_norm": 0.28897902369499207, "learning_rate": 4.12375e-05, "loss": 0.0755, "step": 3300 }, { "epoch": 0.06614443867151593, "grad_norm": 0.25586679577827454, "learning_rate": 4.13625e-05, "loss": 0.0541, "step": 3310 }, { "epoch": 0.06634427081251748, "grad_norm": 0.2735517919063568, "learning_rate": 4.14875e-05, "loss": 0.0532, "step": 3320 }, { "epoch": 0.06654410295351905, "grad_norm": 0.21929825842380524, "learning_rate": 4.16125e-05, "loss": 0.0625, "step": 3330 }, { "epoch": 0.0667439350945206, "grad_norm": 0.24160662293434143, "learning_rate": 4.1737500000000005e-05, "loss": 0.0588, "step": 3340 }, { "epoch": 0.06694376723552216, "grad_norm": 0.24432601034641266, "learning_rate": 4.18625e-05, "loss": 0.0643, "step": 3350 }, { "epoch": 0.06714359937652371, "grad_norm": 0.2607865035533905, "learning_rate": 4.1987500000000004e-05, "loss": 0.0678, "step": 3360 }, { "epoch": 0.06734343151752528, "grad_norm": 0.28801044821739197, "learning_rate": 4.211250000000001e-05, "loss": 0.0701, "step": 3370 }, { "epoch": 0.06754326365852684, "grad_norm": 0.31158071756362915, "learning_rate": 4.2237500000000003e-05, "loss": 0.0546, "step": 3380 }, { "epoch": 0.06774309579952839, "grad_norm": 0.257463276386261, "learning_rate": 4.23625e-05, "loss": 0.0608, "step": 3390 }, { "epoch": 0.06794292794052996, "grad_norm": 0.28090038895606995, "learning_rate": 4.24875e-05, "loss": 0.0606, "step": 3400 }, { "epoch": 0.06814276008153151, "grad_norm": 0.27617132663726807, "learning_rate": 4.26125e-05, "loss": 0.0566, "step": 3410 }, { "epoch": 0.06834259222253307, "grad_norm": 0.28624555468559265, "learning_rate": 4.27375e-05, "loss": 0.0681, "step": 3420 }, { "epoch": 0.06854242436353464, "grad_norm": 0.3036448359489441, "learning_rate": 4.28625e-05, "loss": 0.0582, "step": 3430 }, { "epoch": 0.06874225650453619, "grad_norm": 0.2510046660900116, "learning_rate": 4.29875e-05, "loss": 0.073, "step": 3440 }, { "epoch": 0.06894208864553775, "grad_norm": 0.24962805211544037, "learning_rate": 4.31125e-05, "loss": 0.0649, "step": 3450 }, { "epoch": 0.0691419207865393, "grad_norm": 0.23989474773406982, "learning_rate": 4.32375e-05, "loss": 0.0608, "step": 3460 }, { "epoch": 0.06934175292754087, "grad_norm": 0.21790838241577148, "learning_rate": 4.3362500000000004e-05, "loss": 0.0597, "step": 3470 }, { "epoch": 0.06954158506854242, "grad_norm": 0.24025754630565643, "learning_rate": 4.34875e-05, "loss": 0.0559, "step": 3480 }, { "epoch": 0.06974141720954398, "grad_norm": 0.25276944041252136, "learning_rate": 4.36125e-05, "loss": 0.0663, "step": 3490 }, { "epoch": 0.06994124935054555, "grad_norm": 0.230938121676445, "learning_rate": 4.3737500000000006e-05, "loss": 0.0574, "step": 3500 }, { "epoch": 0.0701410814915471, "grad_norm": 0.2765932083129883, "learning_rate": 4.38625e-05, "loss": 0.062, "step": 3510 }, { "epoch": 0.07034091363254866, "grad_norm": 0.30772513151168823, "learning_rate": 4.3987500000000005e-05, "loss": 0.0673, "step": 3520 }, { "epoch": 0.07054074577355021, "grad_norm": 0.24803155660629272, "learning_rate": 4.41125e-05, "loss": 0.0547, "step": 3530 }, { "epoch": 0.07074057791455178, "grad_norm": 0.26294955611228943, "learning_rate": 4.4237500000000005e-05, "loss": 0.0693, "step": 3540 }, { "epoch": 0.07094041005555334, "grad_norm": 0.21638350188732147, "learning_rate": 4.43625e-05, "loss": 0.0606, "step": 3550 }, { "epoch": 0.07114024219655489, "grad_norm": 0.25924232602119446, "learning_rate": 4.4487500000000004e-05, "loss": 0.0612, "step": 3560 }, { "epoch": 0.07134007433755646, "grad_norm": 0.28526753187179565, "learning_rate": 4.46125e-05, "loss": 0.0654, "step": 3570 }, { "epoch": 0.07153990647855801, "grad_norm": 0.1928544044494629, "learning_rate": 4.47375e-05, "loss": 0.0733, "step": 3580 }, { "epoch": 0.07173973861955957, "grad_norm": 0.2379526048898697, "learning_rate": 4.48625e-05, "loss": 0.0632, "step": 3590 }, { "epoch": 0.07193957076056112, "grad_norm": 0.25399962067604065, "learning_rate": 4.49875e-05, "loss": 0.0633, "step": 3600 }, { "epoch": 0.07213940290156269, "grad_norm": 0.22294245660305023, "learning_rate": 4.51125e-05, "loss": 0.0587, "step": 3610 }, { "epoch": 0.07233923504256425, "grad_norm": 0.2395600825548172, "learning_rate": 4.52375e-05, "loss": 0.0587, "step": 3620 }, { "epoch": 0.0725390671835658, "grad_norm": 0.19705495238304138, "learning_rate": 4.5362500000000005e-05, "loss": 0.0595, "step": 3630 }, { "epoch": 0.07273889932456737, "grad_norm": 0.25795474648475647, "learning_rate": 4.54875e-05, "loss": 0.0551, "step": 3640 }, { "epoch": 0.07293873146556892, "grad_norm": 0.24291297793388367, "learning_rate": 4.5612500000000004e-05, "loss": 0.0624, "step": 3650 }, { "epoch": 0.07313856360657048, "grad_norm": 0.16901151835918427, "learning_rate": 4.57375e-05, "loss": 0.0585, "step": 3660 }, { "epoch": 0.07333839574757203, "grad_norm": 0.23553192615509033, "learning_rate": 4.5862500000000003e-05, "loss": 0.0589, "step": 3670 }, { "epoch": 0.0735382278885736, "grad_norm": 0.26376548409461975, "learning_rate": 4.59875e-05, "loss": 0.0568, "step": 3680 }, { "epoch": 0.07373806002957516, "grad_norm": 0.20537088811397552, "learning_rate": 4.61125e-05, "loss": 0.0591, "step": 3690 }, { "epoch": 0.07393789217057671, "grad_norm": 0.250064492225647, "learning_rate": 4.62375e-05, "loss": 0.0603, "step": 3700 }, { "epoch": 0.07413772431157828, "grad_norm": 0.19677814841270447, "learning_rate": 4.63625e-05, "loss": 0.0654, "step": 3710 }, { "epoch": 0.07433755645257983, "grad_norm": 0.4006437063217163, "learning_rate": 4.64875e-05, "loss": 0.066, "step": 3720 }, { "epoch": 0.07453738859358139, "grad_norm": 0.20792977511882782, "learning_rate": 4.66125e-05, "loss": 0.0594, "step": 3730 }, { "epoch": 0.07473722073458296, "grad_norm": 0.32822057604789734, "learning_rate": 4.67375e-05, "loss": 0.0712, "step": 3740 }, { "epoch": 0.07493705287558451, "grad_norm": 0.273305207490921, "learning_rate": 4.68625e-05, "loss": 0.0626, "step": 3750 }, { "epoch": 0.07513688501658607, "grad_norm": 0.24918925762176514, "learning_rate": 4.6987500000000004e-05, "loss": 0.0553, "step": 3760 }, { "epoch": 0.07533671715758762, "grad_norm": 0.2646113336086273, "learning_rate": 4.711250000000001e-05, "loss": 0.0525, "step": 3770 }, { "epoch": 0.07553654929858919, "grad_norm": 0.23115494847297668, "learning_rate": 4.72375e-05, "loss": 0.0565, "step": 3780 }, { "epoch": 0.07573638143959074, "grad_norm": 0.28061869740486145, "learning_rate": 4.7362500000000006e-05, "loss": 0.067, "step": 3790 }, { "epoch": 0.0759362135805923, "grad_norm": 0.23917809128761292, "learning_rate": 4.74875e-05, "loss": 0.0692, "step": 3800 }, { "epoch": 0.07613604572159387, "grad_norm": 0.2700828015804291, "learning_rate": 4.7612500000000005e-05, "loss": 0.0627, "step": 3810 }, { "epoch": 0.07633587786259542, "grad_norm": 0.3137759268283844, "learning_rate": 4.77375e-05, "loss": 0.0665, "step": 3820 }, { "epoch": 0.07653571000359698, "grad_norm": 0.3078884184360504, "learning_rate": 4.7862500000000005e-05, "loss": 0.0567, "step": 3830 }, { "epoch": 0.07673554214459853, "grad_norm": 0.1781153678894043, "learning_rate": 4.79875e-05, "loss": 0.0657, "step": 3840 }, { "epoch": 0.0769353742856001, "grad_norm": 0.2062133550643921, "learning_rate": 4.8112500000000004e-05, "loss": 0.0572, "step": 3850 }, { "epoch": 0.07713520642660165, "grad_norm": 0.2542942762374878, "learning_rate": 4.82375e-05, "loss": 0.0683, "step": 3860 }, { "epoch": 0.07733503856760321, "grad_norm": 0.2142331749200821, "learning_rate": 4.83625e-05, "loss": 0.064, "step": 3870 }, { "epoch": 0.07753487070860478, "grad_norm": 0.25435805320739746, "learning_rate": 4.84875e-05, "loss": 0.0591, "step": 3880 }, { "epoch": 0.07773470284960633, "grad_norm": 0.2686748206615448, "learning_rate": 4.86125e-05, "loss": 0.0643, "step": 3890 }, { "epoch": 0.07793453499060789, "grad_norm": 0.20022305846214294, "learning_rate": 4.8737500000000006e-05, "loss": 0.0596, "step": 3900 }, { "epoch": 0.07813436713160944, "grad_norm": 0.18685382604599, "learning_rate": 4.88625e-05, "loss": 0.0516, "step": 3910 }, { "epoch": 0.07833419927261101, "grad_norm": 0.2197023183107376, "learning_rate": 4.8987500000000005e-05, "loss": 0.0541, "step": 3920 }, { "epoch": 0.07853403141361257, "grad_norm": 0.2600826621055603, "learning_rate": 4.91125e-05, "loss": 0.0627, "step": 3930 }, { "epoch": 0.07873386355461412, "grad_norm": 0.22448399662971497, "learning_rate": 4.9237500000000004e-05, "loss": 0.0688, "step": 3940 }, { "epoch": 0.07893369569561569, "grad_norm": 0.21657255291938782, "learning_rate": 4.93625e-05, "loss": 0.0604, "step": 3950 }, { "epoch": 0.07913352783661724, "grad_norm": 0.23505795001983643, "learning_rate": 4.9487500000000003e-05, "loss": 0.0738, "step": 3960 }, { "epoch": 0.0793333599776188, "grad_norm": 0.21367786824703217, "learning_rate": 4.96125e-05, "loss": 0.0613, "step": 3970 }, { "epoch": 0.07953319211862035, "grad_norm": 0.21090081334114075, "learning_rate": 4.97375e-05, "loss": 0.0556, "step": 3980 }, { "epoch": 0.07973302425962192, "grad_norm": 0.2002357393503189, "learning_rate": 4.98625e-05, "loss": 0.0562, "step": 3990 }, { "epoch": 0.07993285640062348, "grad_norm": 0.27943339943885803, "learning_rate": 4.99875e-05, "loss": 0.0539, "step": 4000 }, { "epoch": 0.08013268854162503, "grad_norm": 0.22452600300312042, "learning_rate": 5.0112500000000005e-05, "loss": 0.0564, "step": 4010 }, { "epoch": 0.0803325206826266, "grad_norm": 0.2814435362815857, "learning_rate": 5.02375e-05, "loss": 0.056, "step": 4020 }, { "epoch": 0.08053235282362815, "grad_norm": 0.21139413118362427, "learning_rate": 5.0362500000000004e-05, "loss": 0.058, "step": 4030 }, { "epoch": 0.08073218496462971, "grad_norm": 0.28315162658691406, "learning_rate": 5.04875e-05, "loss": 0.0651, "step": 4040 }, { "epoch": 0.08093201710563128, "grad_norm": 0.23275226354599, "learning_rate": 5.061250000000001e-05, "loss": 0.0629, "step": 4050 }, { "epoch": 0.08113184924663283, "grad_norm": 0.27072587609291077, "learning_rate": 5.073750000000001e-05, "loss": 0.0605, "step": 4060 }, { "epoch": 0.08133168138763439, "grad_norm": 0.24219609797000885, "learning_rate": 5.08625e-05, "loss": 0.071, "step": 4070 }, { "epoch": 0.08153151352863594, "grad_norm": 0.23353959619998932, "learning_rate": 5.09875e-05, "loss": 0.0614, "step": 4080 }, { "epoch": 0.08173134566963751, "grad_norm": 0.21632789075374603, "learning_rate": 5.111250000000001e-05, "loss": 0.0616, "step": 4090 }, { "epoch": 0.08193117781063906, "grad_norm": 0.18168872594833374, "learning_rate": 5.1237500000000005e-05, "loss": 0.0583, "step": 4100 }, { "epoch": 0.08213100995164062, "grad_norm": 0.2016826570034027, "learning_rate": 5.13625e-05, "loss": 0.0542, "step": 4110 }, { "epoch": 0.08233084209264219, "grad_norm": 0.19089116156101227, "learning_rate": 5.14875e-05, "loss": 0.0587, "step": 4120 }, { "epoch": 0.08253067423364374, "grad_norm": 0.24411702156066895, "learning_rate": 5.1612499999999994e-05, "loss": 0.0534, "step": 4130 }, { "epoch": 0.0827305063746453, "grad_norm": 0.24966105818748474, "learning_rate": 5.1737500000000004e-05, "loss": 0.0616, "step": 4140 }, { "epoch": 0.08293033851564685, "grad_norm": 0.2380944788455963, "learning_rate": 5.18625e-05, "loss": 0.0571, "step": 4150 }, { "epoch": 0.08313017065664842, "grad_norm": 0.2035621553659439, "learning_rate": 5.19875e-05, "loss": 0.0493, "step": 4160 }, { "epoch": 0.08333000279764997, "grad_norm": 0.15970060229301453, "learning_rate": 5.21125e-05, "loss": 0.0533, "step": 4170 }, { "epoch": 0.08352983493865153, "grad_norm": 0.2250191569328308, "learning_rate": 5.22375e-05, "loss": 0.0589, "step": 4180 }, { "epoch": 0.0837296670796531, "grad_norm": 0.18684576451778412, "learning_rate": 5.2362500000000006e-05, "loss": 0.0575, "step": 4190 }, { "epoch": 0.08392949922065465, "grad_norm": 0.21406331658363342, "learning_rate": 5.24875e-05, "loss": 0.0509, "step": 4200 }, { "epoch": 0.0841293313616562, "grad_norm": 0.27556589245796204, "learning_rate": 5.26125e-05, "loss": 0.0547, "step": 4210 }, { "epoch": 0.08432916350265776, "grad_norm": 0.20479081571102142, "learning_rate": 5.273750000000001e-05, "loss": 0.0575, "step": 4220 }, { "epoch": 0.08452899564365933, "grad_norm": 0.19278763234615326, "learning_rate": 5.2862500000000004e-05, "loss": 0.0554, "step": 4230 }, { "epoch": 0.08472882778466088, "grad_norm": 0.2479635328054428, "learning_rate": 5.29875e-05, "loss": 0.0529, "step": 4240 }, { "epoch": 0.08492865992566244, "grad_norm": 0.1631295084953308, "learning_rate": 5.31125e-05, "loss": 0.0645, "step": 4250 }, { "epoch": 0.08512849206666401, "grad_norm": 0.24994874000549316, "learning_rate": 5.3237500000000007e-05, "loss": 0.0539, "step": 4260 }, { "epoch": 0.08532832420766556, "grad_norm": 0.2473553568124771, "learning_rate": 5.33625e-05, "loss": 0.0609, "step": 4270 }, { "epoch": 0.08552815634866712, "grad_norm": 0.18286418914794922, "learning_rate": 5.34875e-05, "loss": 0.0592, "step": 4280 }, { "epoch": 0.08572798848966867, "grad_norm": 0.23965410888195038, "learning_rate": 5.36125e-05, "loss": 0.0779, "step": 4290 }, { "epoch": 0.08592782063067024, "grad_norm": 0.2228589504957199, "learning_rate": 5.3737500000000005e-05, "loss": 0.0526, "step": 4300 }, { "epoch": 0.0861276527716718, "grad_norm": 0.23449690639972687, "learning_rate": 5.38625e-05, "loss": 0.0566, "step": 4310 }, { "epoch": 0.08632748491267335, "grad_norm": 0.20783111453056335, "learning_rate": 5.3987500000000004e-05, "loss": 0.0562, "step": 4320 }, { "epoch": 0.08652731705367492, "grad_norm": 0.1916452944278717, "learning_rate": 5.41125e-05, "loss": 0.0581, "step": 4330 }, { "epoch": 0.08672714919467647, "grad_norm": 0.211119145154953, "learning_rate": 5.423750000000001e-05, "loss": 0.0483, "step": 4340 }, { "epoch": 0.08692698133567803, "grad_norm": 0.18955092132091522, "learning_rate": 5.436250000000001e-05, "loss": 0.0508, "step": 4350 }, { "epoch": 0.0871268134766796, "grad_norm": 0.2395506054162979, "learning_rate": 5.44875e-05, "loss": 0.0725, "step": 4360 }, { "epoch": 0.08732664561768115, "grad_norm": 0.2466920018196106, "learning_rate": 5.46125e-05, "loss": 0.0503, "step": 4370 }, { "epoch": 0.0875264777586827, "grad_norm": 0.2913496196269989, "learning_rate": 5.4737499999999996e-05, "loss": 0.0609, "step": 4380 }, { "epoch": 0.08772630989968426, "grad_norm": 0.15145660936832428, "learning_rate": 5.4862500000000005e-05, "loss": 0.06, "step": 4390 }, { "epoch": 0.08792614204068583, "grad_norm": 0.19073045253753662, "learning_rate": 5.49875e-05, "loss": 0.0479, "step": 4400 }, { "epoch": 0.08812597418168738, "grad_norm": 0.20358885824680328, "learning_rate": 5.51125e-05, "loss": 0.056, "step": 4410 }, { "epoch": 0.08832580632268894, "grad_norm": 0.2043401002883911, "learning_rate": 5.52375e-05, "loss": 0.0577, "step": 4420 }, { "epoch": 0.0885256384636905, "grad_norm": 0.2563624978065491, "learning_rate": 5.5362500000000004e-05, "loss": 0.0524, "step": 4430 }, { "epoch": 0.08872547060469206, "grad_norm": 0.17397241294384003, "learning_rate": 5.54875e-05, "loss": 0.0593, "step": 4440 }, { "epoch": 0.08892530274569362, "grad_norm": 0.1906370222568512, "learning_rate": 5.56125e-05, "loss": 0.0575, "step": 4450 }, { "epoch": 0.08912513488669517, "grad_norm": 0.3064711391925812, "learning_rate": 5.57375e-05, "loss": 0.063, "step": 4460 }, { "epoch": 0.08932496702769674, "grad_norm": 0.2520204186439514, "learning_rate": 5.586250000000001e-05, "loss": 0.0585, "step": 4470 }, { "epoch": 0.0895247991686983, "grad_norm": 0.2008991539478302, "learning_rate": 5.5987500000000006e-05, "loss": 0.0484, "step": 4480 }, { "epoch": 0.08972463130969985, "grad_norm": 0.3014887571334839, "learning_rate": 5.61125e-05, "loss": 0.0534, "step": 4490 }, { "epoch": 0.08992446345070142, "grad_norm": 0.23187607526779175, "learning_rate": 5.62375e-05, "loss": 0.072, "step": 4500 }, { "epoch": 0.09012429559170297, "grad_norm": 0.19967223703861237, "learning_rate": 5.636250000000001e-05, "loss": 0.0522, "step": 4510 }, { "epoch": 0.09032412773270453, "grad_norm": 0.20162594318389893, "learning_rate": 5.6487500000000004e-05, "loss": 0.067, "step": 4520 }, { "epoch": 0.09052395987370608, "grad_norm": 0.22555425763130188, "learning_rate": 5.66125e-05, "loss": 0.0586, "step": 4530 }, { "epoch": 0.09072379201470765, "grad_norm": 0.2747099697589874, "learning_rate": 5.67375e-05, "loss": 0.0521, "step": 4540 }, { "epoch": 0.0909236241557092, "grad_norm": 0.21576732397079468, "learning_rate": 5.6862500000000007e-05, "loss": 0.0661, "step": 4550 }, { "epoch": 0.09112345629671076, "grad_norm": 0.2096843421459198, "learning_rate": 5.69875e-05, "loss": 0.0626, "step": 4560 }, { "epoch": 0.09132328843771233, "grad_norm": 0.20207010209560394, "learning_rate": 5.71125e-05, "loss": 0.0548, "step": 4570 }, { "epoch": 0.09152312057871388, "grad_norm": 0.20086048543453217, "learning_rate": 5.72375e-05, "loss": 0.0598, "step": 4580 }, { "epoch": 0.09172295271971544, "grad_norm": 0.31422820687294006, "learning_rate": 5.7362500000000005e-05, "loss": 0.0512, "step": 4590 }, { "epoch": 0.09192278486071699, "grad_norm": 0.2237008959054947, "learning_rate": 5.748750000000001e-05, "loss": 0.0517, "step": 4600 }, { "epoch": 0.09212261700171856, "grad_norm": 0.19665193557739258, "learning_rate": 5.7612500000000004e-05, "loss": 0.0653, "step": 4610 }, { "epoch": 0.09232244914272011, "grad_norm": 0.2106550931930542, "learning_rate": 5.77375e-05, "loss": 0.0617, "step": 4620 }, { "epoch": 0.09252228128372167, "grad_norm": 0.23704221844673157, "learning_rate": 5.78625e-05, "loss": 0.0502, "step": 4630 }, { "epoch": 0.09272211342472324, "grad_norm": 0.21553196012973785, "learning_rate": 5.798750000000001e-05, "loss": 0.0592, "step": 4640 }, { "epoch": 0.09292194556572479, "grad_norm": 0.22553221881389618, "learning_rate": 5.81125e-05, "loss": 0.0665, "step": 4650 }, { "epoch": 0.09312177770672635, "grad_norm": 0.30348673462867737, "learning_rate": 5.82375e-05, "loss": 0.0638, "step": 4660 }, { "epoch": 0.09332160984772792, "grad_norm": 0.23536156117916107, "learning_rate": 5.8362499999999996e-05, "loss": 0.0616, "step": 4670 }, { "epoch": 0.09352144198872947, "grad_norm": 0.2798740267753601, "learning_rate": 5.8487500000000005e-05, "loss": 0.0526, "step": 4680 }, { "epoch": 0.09372127412973102, "grad_norm": 0.24222925305366516, "learning_rate": 5.86125e-05, "loss": 0.0621, "step": 4690 }, { "epoch": 0.09392110627073258, "grad_norm": 0.279509037733078, "learning_rate": 5.87375e-05, "loss": 0.0622, "step": 4700 }, { "epoch": 0.09412093841173415, "grad_norm": 0.20593494176864624, "learning_rate": 5.88625e-05, "loss": 0.0565, "step": 4710 }, { "epoch": 0.0943207705527357, "grad_norm": 0.17936372756958008, "learning_rate": 5.8987500000000004e-05, "loss": 0.0522, "step": 4720 }, { "epoch": 0.09452060269373726, "grad_norm": 0.16832736134529114, "learning_rate": 5.911250000000001e-05, "loss": 0.0494, "step": 4730 }, { "epoch": 0.09472043483473883, "grad_norm": 0.17063096165657043, "learning_rate": 5.92375e-05, "loss": 0.0613, "step": 4740 }, { "epoch": 0.09492026697574038, "grad_norm": 0.29038694500923157, "learning_rate": 5.93625e-05, "loss": 0.0522, "step": 4750 }, { "epoch": 0.09512009911674194, "grad_norm": 0.28742438554763794, "learning_rate": 5.948750000000001e-05, "loss": 0.0548, "step": 4760 }, { "epoch": 0.09531993125774349, "grad_norm": 0.19352874159812927, "learning_rate": 5.9612500000000006e-05, "loss": 0.0626, "step": 4770 }, { "epoch": 0.09551976339874506, "grad_norm": 0.2205471396446228, "learning_rate": 5.97375e-05, "loss": 0.0613, "step": 4780 }, { "epoch": 0.09571959553974661, "grad_norm": 0.22272101044654846, "learning_rate": 5.98625e-05, "loss": 0.0654, "step": 4790 }, { "epoch": 0.09591942768074817, "grad_norm": 0.1884399801492691, "learning_rate": 5.998750000000001e-05, "loss": 0.0546, "step": 4800 }, { "epoch": 0.09611925982174974, "grad_norm": 0.17296160757541656, "learning_rate": 6.0112500000000004e-05, "loss": 0.0574, "step": 4810 }, { "epoch": 0.09631909196275129, "grad_norm": 0.21557675302028656, "learning_rate": 6.02375e-05, "loss": 0.0516, "step": 4820 }, { "epoch": 0.09651892410375285, "grad_norm": 0.19874230027198792, "learning_rate": 6.03625e-05, "loss": 0.0555, "step": 4830 }, { "epoch": 0.0967187562447544, "grad_norm": 0.16109557449817657, "learning_rate": 6.0487500000000007e-05, "loss": 0.0573, "step": 4840 }, { "epoch": 0.09691858838575597, "grad_norm": 0.20466147363185883, "learning_rate": 6.06125e-05, "loss": 0.0551, "step": 4850 }, { "epoch": 0.09711842052675752, "grad_norm": 0.22452139854431152, "learning_rate": 6.0737500000000006e-05, "loss": 0.0568, "step": 4860 }, { "epoch": 0.09731825266775908, "grad_norm": 0.21977722644805908, "learning_rate": 6.08625e-05, "loss": 0.0641, "step": 4870 }, { "epoch": 0.09751808480876065, "grad_norm": 0.17382165789604187, "learning_rate": 6.09875e-05, "loss": 0.057, "step": 4880 }, { "epoch": 0.0977179169497622, "grad_norm": 0.17230820655822754, "learning_rate": 6.11125e-05, "loss": 0.0587, "step": 4890 }, { "epoch": 0.09791774909076376, "grad_norm": 0.2682265043258667, "learning_rate": 6.12375e-05, "loss": 0.0526, "step": 4900 }, { "epoch": 0.09811758123176531, "grad_norm": 0.2410694658756256, "learning_rate": 6.13625e-05, "loss": 0.0604, "step": 4910 }, { "epoch": 0.09831741337276688, "grad_norm": 0.19021981954574585, "learning_rate": 6.14875e-05, "loss": 0.0628, "step": 4920 }, { "epoch": 0.09851724551376843, "grad_norm": 0.22289547324180603, "learning_rate": 6.16125e-05, "loss": 0.0541, "step": 4930 }, { "epoch": 0.09871707765476999, "grad_norm": 0.2609409987926483, "learning_rate": 6.17375e-05, "loss": 0.0596, "step": 4940 }, { "epoch": 0.09891690979577156, "grad_norm": 0.20639578998088837, "learning_rate": 6.18625e-05, "loss": 0.0523, "step": 4950 }, { "epoch": 0.09911674193677311, "grad_norm": 0.1688123643398285, "learning_rate": 6.19875e-05, "loss": 0.0616, "step": 4960 }, { "epoch": 0.09931657407777467, "grad_norm": 0.22197118401527405, "learning_rate": 6.211250000000001e-05, "loss": 0.0534, "step": 4970 }, { "epoch": 0.09951640621877622, "grad_norm": 0.17071951925754547, "learning_rate": 6.22375e-05, "loss": 0.0483, "step": 4980 }, { "epoch": 0.09971623835977779, "grad_norm": 0.1743047535419464, "learning_rate": 6.23625e-05, "loss": 0.0554, "step": 4990 }, { "epoch": 0.09991607050077934, "grad_norm": 0.2786470949649811, "learning_rate": 6.24875e-05, "loss": 0.0655, "step": 5000 }, { "epoch": 0.1001159026417809, "grad_norm": 0.2307179719209671, "learning_rate": 6.261250000000001e-05, "loss": 0.0588, "step": 5010 }, { "epoch": 0.10031573478278247, "grad_norm": 0.24078381061553955, "learning_rate": 6.27375e-05, "loss": 0.0586, "step": 5020 }, { "epoch": 0.10051556692378402, "grad_norm": 0.2106763869524002, "learning_rate": 6.28625e-05, "loss": 0.0583, "step": 5030 }, { "epoch": 0.10071539906478558, "grad_norm": 0.1949121356010437, "learning_rate": 6.298749999999999e-05, "loss": 0.0514, "step": 5040 }, { "epoch": 0.10091523120578715, "grad_norm": 0.21185027062892914, "learning_rate": 6.311250000000001e-05, "loss": 0.0587, "step": 5050 }, { "epoch": 0.1011150633467887, "grad_norm": 0.21182283759117126, "learning_rate": 6.32375e-05, "loss": 0.0473, "step": 5060 }, { "epoch": 0.10131489548779025, "grad_norm": 0.19724377989768982, "learning_rate": 6.33625e-05, "loss": 0.0574, "step": 5070 }, { "epoch": 0.10151472762879181, "grad_norm": 0.24992208182811737, "learning_rate": 6.34875e-05, "loss": 0.0629, "step": 5080 }, { "epoch": 0.10171455976979338, "grad_norm": 0.2299012988805771, "learning_rate": 6.361250000000001e-05, "loss": 0.0634, "step": 5090 }, { "epoch": 0.10191439191079493, "grad_norm": 0.23300699889659882, "learning_rate": 6.373750000000001e-05, "loss": 0.0664, "step": 5100 }, { "epoch": 0.10211422405179649, "grad_norm": 0.20622767508029938, "learning_rate": 6.38625e-05, "loss": 0.056, "step": 5110 }, { "epoch": 0.10231405619279806, "grad_norm": 0.2400280237197876, "learning_rate": 6.39875e-05, "loss": 0.0576, "step": 5120 }, { "epoch": 0.10251388833379961, "grad_norm": 0.22564846277236938, "learning_rate": 6.411249999999999e-05, "loss": 0.0526, "step": 5130 }, { "epoch": 0.10271372047480117, "grad_norm": 0.20431192219257355, "learning_rate": 6.423750000000001e-05, "loss": 0.0645, "step": 5140 }, { "epoch": 0.10291355261580272, "grad_norm": 0.2132871299982071, "learning_rate": 6.43625e-05, "loss": 0.0613, "step": 5150 }, { "epoch": 0.10311338475680429, "grad_norm": 0.18560709059238434, "learning_rate": 6.44875e-05, "loss": 0.0566, "step": 5160 }, { "epoch": 0.10331321689780584, "grad_norm": 0.22709375619888306, "learning_rate": 6.461249999999999e-05, "loss": 0.0592, "step": 5170 }, { "epoch": 0.1035130490388074, "grad_norm": 0.21715998649597168, "learning_rate": 6.473750000000001e-05, "loss": 0.0631, "step": 5180 }, { "epoch": 0.10371288117980897, "grad_norm": 0.2368939220905304, "learning_rate": 6.48625e-05, "loss": 0.056, "step": 5190 }, { "epoch": 0.10391271332081052, "grad_norm": 0.22972799837589264, "learning_rate": 6.49875e-05, "loss": 0.0583, "step": 5200 }, { "epoch": 0.10411254546181208, "grad_norm": 0.1927158683538437, "learning_rate": 6.51125e-05, "loss": 0.0581, "step": 5210 }, { "epoch": 0.10431237760281363, "grad_norm": 0.20396265387535095, "learning_rate": 6.523750000000001e-05, "loss": 0.0512, "step": 5220 }, { "epoch": 0.1045122097438152, "grad_norm": 0.23116706311702728, "learning_rate": 6.536250000000001e-05, "loss": 0.0601, "step": 5230 }, { "epoch": 0.10471204188481675, "grad_norm": 0.2270064800977707, "learning_rate": 6.54875e-05, "loss": 0.0548, "step": 5240 }, { "epoch": 0.10491187402581831, "grad_norm": 0.1615624874830246, "learning_rate": 6.56125e-05, "loss": 0.0507, "step": 5250 }, { "epoch": 0.10511170616681988, "grad_norm": 0.28583648800849915, "learning_rate": 6.57375e-05, "loss": 0.0566, "step": 5260 }, { "epoch": 0.10531153830782143, "grad_norm": 0.3069015443325043, "learning_rate": 6.586250000000001e-05, "loss": 0.0547, "step": 5270 }, { "epoch": 0.10551137044882299, "grad_norm": 0.1783255785703659, "learning_rate": 6.59875e-05, "loss": 0.0449, "step": 5280 }, { "epoch": 0.10571120258982454, "grad_norm": 0.2260257750749588, "learning_rate": 6.61125e-05, "loss": 0.0522, "step": 5290 }, { "epoch": 0.10591103473082611, "grad_norm": 0.20140275359153748, "learning_rate": 6.62375e-05, "loss": 0.0634, "step": 5300 }, { "epoch": 0.10611086687182766, "grad_norm": 0.18684574961662292, "learning_rate": 6.636250000000001e-05, "loss": 0.0747, "step": 5310 }, { "epoch": 0.10631069901282922, "grad_norm": 0.1917361170053482, "learning_rate": 6.64875e-05, "loss": 0.058, "step": 5320 }, { "epoch": 0.10651053115383079, "grad_norm": 0.19623611867427826, "learning_rate": 6.66125e-05, "loss": 0.0642, "step": 5330 }, { "epoch": 0.10671036329483234, "grad_norm": 0.22550621628761292, "learning_rate": 6.67375e-05, "loss": 0.0541, "step": 5340 }, { "epoch": 0.1069101954358339, "grad_norm": 0.1878121942281723, "learning_rate": 6.68625e-05, "loss": 0.0618, "step": 5350 }, { "epoch": 0.10711002757683546, "grad_norm": 0.2217540591955185, "learning_rate": 6.698750000000001e-05, "loss": 0.0586, "step": 5360 }, { "epoch": 0.10730985971783702, "grad_norm": 0.19655349850654602, "learning_rate": 6.71125e-05, "loss": 0.047, "step": 5370 }, { "epoch": 0.10750969185883857, "grad_norm": 0.1808442920446396, "learning_rate": 6.72375e-05, "loss": 0.064, "step": 5380 }, { "epoch": 0.10770952399984013, "grad_norm": 0.22986148297786713, "learning_rate": 6.73625e-05, "loss": 0.061, "step": 5390 }, { "epoch": 0.1079093561408417, "grad_norm": 0.1831485480070114, "learning_rate": 6.748750000000001e-05, "loss": 0.0615, "step": 5400 }, { "epoch": 0.10810918828184325, "grad_norm": 0.21734994649887085, "learning_rate": 6.76125e-05, "loss": 0.0564, "step": 5410 }, { "epoch": 0.1083090204228448, "grad_norm": 0.2127661556005478, "learning_rate": 6.77375e-05, "loss": 0.0595, "step": 5420 }, { "epoch": 0.10850885256384638, "grad_norm": 0.2393825352191925, "learning_rate": 6.78625e-05, "loss": 0.055, "step": 5430 }, { "epoch": 0.10870868470484793, "grad_norm": 0.20202945172786713, "learning_rate": 6.79875e-05, "loss": 0.0595, "step": 5440 }, { "epoch": 0.10890851684584948, "grad_norm": 0.20080091059207916, "learning_rate": 6.81125e-05, "loss": 0.0602, "step": 5450 }, { "epoch": 0.10910834898685104, "grad_norm": 0.23956596851348877, "learning_rate": 6.82375e-05, "loss": 0.055, "step": 5460 }, { "epoch": 0.10930818112785261, "grad_norm": 0.21835407614707947, "learning_rate": 6.83625e-05, "loss": 0.0675, "step": 5470 }, { "epoch": 0.10950801326885416, "grad_norm": 0.27322110533714294, "learning_rate": 6.84875e-05, "loss": 0.0508, "step": 5480 }, { "epoch": 0.10970784540985572, "grad_norm": 0.15009534358978271, "learning_rate": 6.861250000000001e-05, "loss": 0.0523, "step": 5490 }, { "epoch": 0.10990767755085729, "grad_norm": 0.2296639084815979, "learning_rate": 6.87375e-05, "loss": 0.0512, "step": 5500 }, { "epoch": 0.11010750969185884, "grad_norm": 0.3099086284637451, "learning_rate": 6.886250000000001e-05, "loss": 0.0538, "step": 5510 }, { "epoch": 0.1103073418328604, "grad_norm": 0.23671956360340118, "learning_rate": 6.89875e-05, "loss": 0.0564, "step": 5520 }, { "epoch": 0.11050717397386195, "grad_norm": 0.19027848541736603, "learning_rate": 6.91125e-05, "loss": 0.0554, "step": 5530 }, { "epoch": 0.11070700611486352, "grad_norm": 0.1910037100315094, "learning_rate": 6.92375e-05, "loss": 0.0541, "step": 5540 }, { "epoch": 0.11090683825586507, "grad_norm": 0.2046436071395874, "learning_rate": 6.936250000000001e-05, "loss": 0.0534, "step": 5550 }, { "epoch": 0.11110667039686663, "grad_norm": 0.23062250018119812, "learning_rate": 6.94875e-05, "loss": 0.0619, "step": 5560 }, { "epoch": 0.1113065025378682, "grad_norm": 0.26018479466438293, "learning_rate": 6.96125e-05, "loss": 0.0553, "step": 5570 }, { "epoch": 0.11150633467886975, "grad_norm": 0.22966767847537994, "learning_rate": 6.97375e-05, "loss": 0.0494, "step": 5580 }, { "epoch": 0.1117061668198713, "grad_norm": 0.2477559745311737, "learning_rate": 6.986250000000001e-05, "loss": 0.0532, "step": 5590 }, { "epoch": 0.11190599896087286, "grad_norm": 0.25172072649002075, "learning_rate": 6.99875e-05, "loss": 0.0623, "step": 5600 }, { "epoch": 0.11210583110187443, "grad_norm": 0.2164984941482544, "learning_rate": 7.01125e-05, "loss": 0.0572, "step": 5610 }, { "epoch": 0.11230566324287598, "grad_norm": 0.3141725957393646, "learning_rate": 7.02375e-05, "loss": 0.0587, "step": 5620 }, { "epoch": 0.11250549538387754, "grad_norm": 0.1798037439584732, "learning_rate": 7.03625e-05, "loss": 0.0448, "step": 5630 }, { "epoch": 0.1127053275248791, "grad_norm": 0.23449409008026123, "learning_rate": 7.048750000000001e-05, "loss": 0.0598, "step": 5640 }, { "epoch": 0.11290515966588066, "grad_norm": 0.21961453557014465, "learning_rate": 7.06125e-05, "loss": 0.0573, "step": 5650 }, { "epoch": 0.11310499180688222, "grad_norm": 0.18123260140419006, "learning_rate": 7.07375e-05, "loss": 0.0541, "step": 5660 }, { "epoch": 0.11330482394788378, "grad_norm": 0.15178316831588745, "learning_rate": 7.08625e-05, "loss": 0.0482, "step": 5670 }, { "epoch": 0.11350465608888534, "grad_norm": 0.20870627462863922, "learning_rate": 7.098750000000001e-05, "loss": 0.0634, "step": 5680 }, { "epoch": 0.1137044882298869, "grad_norm": 0.18195635080337524, "learning_rate": 7.11125e-05, "loss": 0.0478, "step": 5690 }, { "epoch": 0.11390432037088845, "grad_norm": 0.20809337496757507, "learning_rate": 7.12375e-05, "loss": 0.0638, "step": 5700 }, { "epoch": 0.11410415251189002, "grad_norm": 0.19399486482143402, "learning_rate": 7.136249999999999e-05, "loss": 0.0587, "step": 5710 }, { "epoch": 0.11430398465289157, "grad_norm": 0.1742868423461914, "learning_rate": 7.148750000000001e-05, "loss": 0.0513, "step": 5720 }, { "epoch": 0.11450381679389313, "grad_norm": 0.19678671658039093, "learning_rate": 7.16125e-05, "loss": 0.0597, "step": 5730 }, { "epoch": 0.1147036489348947, "grad_norm": 0.1685638576745987, "learning_rate": 7.17375e-05, "loss": 0.0482, "step": 5740 }, { "epoch": 0.11490348107589625, "grad_norm": 0.1927807331085205, "learning_rate": 7.18625e-05, "loss": 0.0559, "step": 5750 }, { "epoch": 0.1151033132168978, "grad_norm": 0.20333480834960938, "learning_rate": 7.198750000000001e-05, "loss": 0.0481, "step": 5760 }, { "epoch": 0.11530314535789936, "grad_norm": 0.20092488825321198, "learning_rate": 7.211250000000001e-05, "loss": 0.0518, "step": 5770 }, { "epoch": 0.11550297749890093, "grad_norm": 0.21369990706443787, "learning_rate": 7.22375e-05, "loss": 0.0586, "step": 5780 }, { "epoch": 0.11570280963990248, "grad_norm": 0.26171594858169556, "learning_rate": 7.23625e-05, "loss": 0.0565, "step": 5790 }, { "epoch": 0.11590264178090404, "grad_norm": 0.2662728726863861, "learning_rate": 7.248750000000001e-05, "loss": 0.056, "step": 5800 }, { "epoch": 0.1161024739219056, "grad_norm": 0.23007787764072418, "learning_rate": 7.261250000000001e-05, "loss": 0.0778, "step": 5810 }, { "epoch": 0.11630230606290716, "grad_norm": 0.20539596676826477, "learning_rate": 7.27375e-05, "loss": 0.0649, "step": 5820 }, { "epoch": 0.11650213820390871, "grad_norm": 0.22103023529052734, "learning_rate": 7.28625e-05, "loss": 0.0595, "step": 5830 }, { "epoch": 0.11670197034491027, "grad_norm": 0.20585504174232483, "learning_rate": 7.29875e-05, "loss": 0.0545, "step": 5840 }, { "epoch": 0.11690180248591184, "grad_norm": 0.23093938827514648, "learning_rate": 7.311250000000001e-05, "loss": 0.0581, "step": 5850 }, { "epoch": 0.11710163462691339, "grad_norm": 0.24229827523231506, "learning_rate": 7.32375e-05, "loss": 0.0579, "step": 5860 }, { "epoch": 0.11730146676791495, "grad_norm": 0.2411719113588333, "learning_rate": 7.33625e-05, "loss": 0.0502, "step": 5870 }, { "epoch": 0.11750129890891652, "grad_norm": 0.18325525522232056, "learning_rate": 7.34875e-05, "loss": 0.0577, "step": 5880 }, { "epoch": 0.11770113104991807, "grad_norm": 0.2192903757095337, "learning_rate": 7.361250000000001e-05, "loss": 0.0551, "step": 5890 }, { "epoch": 0.11790096319091962, "grad_norm": 0.22276031970977783, "learning_rate": 7.373750000000001e-05, "loss": 0.0554, "step": 5900 }, { "epoch": 0.11810079533192118, "grad_norm": 0.22742359340190887, "learning_rate": 7.38625e-05, "loss": 0.0661, "step": 5910 }, { "epoch": 0.11830062747292275, "grad_norm": 0.18121501803398132, "learning_rate": 7.39875e-05, "loss": 0.0618, "step": 5920 }, { "epoch": 0.1185004596139243, "grad_norm": 0.22343532741069794, "learning_rate": 7.41125e-05, "loss": 0.0585, "step": 5930 }, { "epoch": 0.11870029175492586, "grad_norm": 0.2120765596628189, "learning_rate": 7.423750000000001e-05, "loss": 0.0571, "step": 5940 }, { "epoch": 0.11890012389592743, "grad_norm": 0.23644165694713593, "learning_rate": 7.43625e-05, "loss": 0.0557, "step": 5950 }, { "epoch": 0.11909995603692898, "grad_norm": 0.15942950546741486, "learning_rate": 7.44875e-05, "loss": 0.0536, "step": 5960 }, { "epoch": 0.11929978817793054, "grad_norm": 0.2947235405445099, "learning_rate": 7.46125e-05, "loss": 0.0553, "step": 5970 }, { "epoch": 0.1194996203189321, "grad_norm": 0.19956733286380768, "learning_rate": 7.473750000000001e-05, "loss": 0.0467, "step": 5980 }, { "epoch": 0.11969945245993366, "grad_norm": 0.15774857997894287, "learning_rate": 7.48625e-05, "loss": 0.0569, "step": 5990 }, { "epoch": 0.11989928460093521, "grad_norm": 0.21583271026611328, "learning_rate": 7.49875e-05, "loss": 0.0624, "step": 6000 }, { "epoch": 0.12009911674193677, "grad_norm": 0.23062430322170258, "learning_rate": 7.51125e-05, "loss": 0.0553, "step": 6010 }, { "epoch": 0.12029894888293834, "grad_norm": 0.17672981321811676, "learning_rate": 7.52375e-05, "loss": 0.0522, "step": 6020 }, { "epoch": 0.12049878102393989, "grad_norm": 0.18275022506713867, "learning_rate": 7.536250000000001e-05, "loss": 0.0605, "step": 6030 }, { "epoch": 0.12069861316494145, "grad_norm": 0.23651538789272308, "learning_rate": 7.54875e-05, "loss": 0.064, "step": 6040 }, { "epoch": 0.12089844530594301, "grad_norm": 0.2033165544271469, "learning_rate": 7.561250000000002e-05, "loss": 0.058, "step": 6050 }, { "epoch": 0.12109827744694457, "grad_norm": 0.15575814247131348, "learning_rate": 7.57375e-05, "loss": 0.0499, "step": 6060 }, { "epoch": 0.12129810958794612, "grad_norm": 0.23894591629505157, "learning_rate": 7.586250000000001e-05, "loss": 0.0514, "step": 6070 }, { "epoch": 0.12149794172894768, "grad_norm": 0.17901988327503204, "learning_rate": 7.59875e-05, "loss": 0.0476, "step": 6080 }, { "epoch": 0.12169777386994925, "grad_norm": 0.12038496881723404, "learning_rate": 7.611250000000001e-05, "loss": 0.0548, "step": 6090 }, { "epoch": 0.1218976060109508, "grad_norm": 0.22591258585453033, "learning_rate": 7.62375e-05, "loss": 0.0564, "step": 6100 }, { "epoch": 0.12209743815195236, "grad_norm": 0.28335100412368774, "learning_rate": 7.63625e-05, "loss": 0.0615, "step": 6110 }, { "epoch": 0.12229727029295392, "grad_norm": 0.24306590855121613, "learning_rate": 7.64875e-05, "loss": 0.0548, "step": 6120 }, { "epoch": 0.12249710243395548, "grad_norm": 0.29585230350494385, "learning_rate": 7.66125e-05, "loss": 0.0676, "step": 6130 }, { "epoch": 0.12269693457495703, "grad_norm": 0.187435120344162, "learning_rate": 7.67375e-05, "loss": 0.0589, "step": 6140 }, { "epoch": 0.12289676671595859, "grad_norm": 0.14800797402858734, "learning_rate": 7.68625e-05, "loss": 0.054, "step": 6150 }, { "epoch": 0.12309659885696016, "grad_norm": 0.19795610010623932, "learning_rate": 7.698750000000001e-05, "loss": 0.061, "step": 6160 }, { "epoch": 0.12329643099796171, "grad_norm": 0.1691831797361374, "learning_rate": 7.71125e-05, "loss": 0.0569, "step": 6170 }, { "epoch": 0.12349626313896327, "grad_norm": 0.22400186955928802, "learning_rate": 7.72375e-05, "loss": 0.0601, "step": 6180 }, { "epoch": 0.12369609527996484, "grad_norm": 0.20141957700252533, "learning_rate": 7.73625e-05, "loss": 0.0505, "step": 6190 }, { "epoch": 0.12389592742096639, "grad_norm": 0.21205399930477142, "learning_rate": 7.74875e-05, "loss": 0.0651, "step": 6200 }, { "epoch": 0.12409575956196794, "grad_norm": 0.18715211749076843, "learning_rate": 7.76125e-05, "loss": 0.0565, "step": 6210 }, { "epoch": 0.1242955917029695, "grad_norm": 0.17160865664482117, "learning_rate": 7.773750000000001e-05, "loss": 0.0668, "step": 6220 }, { "epoch": 0.12449542384397107, "grad_norm": 0.30469048023223877, "learning_rate": 7.78625e-05, "loss": 0.0563, "step": 6230 }, { "epoch": 0.12469525598497262, "grad_norm": 0.18965043127536774, "learning_rate": 7.79875e-05, "loss": 0.0499, "step": 6240 }, { "epoch": 0.12489508812597418, "grad_norm": 0.18752644956111908, "learning_rate": 7.81125e-05, "loss": 0.0553, "step": 6250 }, { "epoch": 0.12509492026697575, "grad_norm": 0.1701122522354126, "learning_rate": 7.823750000000001e-05, "loss": 0.0636, "step": 6260 }, { "epoch": 0.1252947524079773, "grad_norm": 0.20445045828819275, "learning_rate": 7.83625e-05, "loss": 0.0598, "step": 6270 }, { "epoch": 0.12549458454897885, "grad_norm": 0.22629055380821228, "learning_rate": 7.84875e-05, "loss": 0.056, "step": 6280 }, { "epoch": 0.1256944166899804, "grad_norm": 0.2404785305261612, "learning_rate": 7.86125e-05, "loss": 0.0557, "step": 6290 }, { "epoch": 0.12589424883098196, "grad_norm": 0.24672073125839233, "learning_rate": 7.873750000000001e-05, "loss": 0.0593, "step": 6300 }, { "epoch": 0.12609408097198355, "grad_norm": 0.2290174514055252, "learning_rate": 7.88625e-05, "loss": 0.051, "step": 6310 }, { "epoch": 0.1262939131129851, "grad_norm": 0.1709013730287552, "learning_rate": 7.89875e-05, "loss": 0.0593, "step": 6320 }, { "epoch": 0.12649374525398666, "grad_norm": 0.261339396238327, "learning_rate": 7.91125e-05, "loss": 0.0607, "step": 6330 }, { "epoch": 0.1266935773949882, "grad_norm": 0.22619232535362244, "learning_rate": 7.923750000000001e-05, "loss": 0.0559, "step": 6340 }, { "epoch": 0.12689340953598977, "grad_norm": 0.18344691395759583, "learning_rate": 7.936250000000001e-05, "loss": 0.0545, "step": 6350 }, { "epoch": 0.12709324167699132, "grad_norm": 0.35658055543899536, "learning_rate": 7.94875e-05, "loss": 0.0534, "step": 6360 }, { "epoch": 0.12729307381799287, "grad_norm": 0.2102549821138382, "learning_rate": 7.96125e-05, "loss": 0.057, "step": 6370 }, { "epoch": 0.12749290595899446, "grad_norm": 0.19791094958782196, "learning_rate": 7.97375e-05, "loss": 0.0586, "step": 6380 }, { "epoch": 0.127692738099996, "grad_norm": 0.1907825767993927, "learning_rate": 7.986250000000001e-05, "loss": 0.0654, "step": 6390 }, { "epoch": 0.12789257024099757, "grad_norm": 0.19972316920757294, "learning_rate": 7.99875e-05, "loss": 0.0586, "step": 6400 }, { "epoch": 0.12809240238199912, "grad_norm": 0.22789278626441956, "learning_rate": 8.01125e-05, "loss": 0.0594, "step": 6410 }, { "epoch": 0.12829223452300068, "grad_norm": 0.16059616208076477, "learning_rate": 8.02375e-05, "loss": 0.0533, "step": 6420 }, { "epoch": 0.12849206666400223, "grad_norm": 0.25117790699005127, "learning_rate": 8.036250000000001e-05, "loss": 0.0579, "step": 6430 }, { "epoch": 0.12869189880500378, "grad_norm": 0.17353114485740662, "learning_rate": 8.04875e-05, "loss": 0.0483, "step": 6440 }, { "epoch": 0.12889173094600537, "grad_norm": 0.2372514307498932, "learning_rate": 8.06125e-05, "loss": 0.0638, "step": 6450 }, { "epoch": 0.12909156308700692, "grad_norm": 0.15602000057697296, "learning_rate": 8.07375e-05, "loss": 0.0471, "step": 6460 }, { "epoch": 0.12929139522800848, "grad_norm": 0.16260530054569244, "learning_rate": 8.086250000000001e-05, "loss": 0.0522, "step": 6470 }, { "epoch": 0.12949122736901003, "grad_norm": 0.21298334002494812, "learning_rate": 8.098750000000001e-05, "loss": 0.0452, "step": 6480 }, { "epoch": 0.12969105951001159, "grad_norm": 0.20877021551132202, "learning_rate": 8.11125e-05, "loss": 0.0483, "step": 6490 }, { "epoch": 0.12989089165101314, "grad_norm": 0.19108551740646362, "learning_rate": 8.12375e-05, "loss": 0.0535, "step": 6500 }, { "epoch": 0.1300907237920147, "grad_norm": 0.23005743324756622, "learning_rate": 8.13625e-05, "loss": 0.0673, "step": 6510 }, { "epoch": 0.13029055593301628, "grad_norm": 0.2113576978445053, "learning_rate": 8.148750000000001e-05, "loss": 0.0607, "step": 6520 }, { "epoch": 0.13049038807401783, "grad_norm": 0.19214804470539093, "learning_rate": 8.16125e-05, "loss": 0.0507, "step": 6530 }, { "epoch": 0.1306902202150194, "grad_norm": 0.2057480365037918, "learning_rate": 8.17375e-05, "loss": 0.0642, "step": 6540 }, { "epoch": 0.13089005235602094, "grad_norm": 0.2350068837404251, "learning_rate": 8.18625e-05, "loss": 0.0562, "step": 6550 }, { "epoch": 0.1310898844970225, "grad_norm": 0.21349652111530304, "learning_rate": 8.198750000000001e-05, "loss": 0.0463, "step": 6560 }, { "epoch": 0.13128971663802405, "grad_norm": 0.2410929799079895, "learning_rate": 8.21125e-05, "loss": 0.0662, "step": 6570 }, { "epoch": 0.1314895487790256, "grad_norm": 0.1859789341688156, "learning_rate": 8.22375e-05, "loss": 0.051, "step": 6580 }, { "epoch": 0.1316893809200272, "grad_norm": 0.19334451854228973, "learning_rate": 8.23625e-05, "loss": 0.0519, "step": 6590 }, { "epoch": 0.13188921306102874, "grad_norm": 0.27982497215270996, "learning_rate": 8.24875e-05, "loss": 0.0517, "step": 6600 }, { "epoch": 0.1320890452020303, "grad_norm": 0.222842276096344, "learning_rate": 8.261250000000001e-05, "loss": 0.0431, "step": 6610 }, { "epoch": 0.13228887734303185, "grad_norm": 0.19603389501571655, "learning_rate": 8.27375e-05, "loss": 0.0597, "step": 6620 }, { "epoch": 0.1324887094840334, "grad_norm": 0.15558570623397827, "learning_rate": 8.28625e-05, "loss": 0.0516, "step": 6630 }, { "epoch": 0.13268854162503496, "grad_norm": 0.19242727756500244, "learning_rate": 8.29875e-05, "loss": 0.0493, "step": 6640 }, { "epoch": 0.13288837376603652, "grad_norm": 0.1818762868642807, "learning_rate": 8.311250000000001e-05, "loss": 0.0502, "step": 6650 }, { "epoch": 0.1330882059070381, "grad_norm": 0.225604847073555, "learning_rate": 8.32375e-05, "loss": 0.0662, "step": 6660 }, { "epoch": 0.13328803804803965, "grad_norm": 0.15209795534610748, "learning_rate": 8.33625e-05, "loss": 0.056, "step": 6670 }, { "epoch": 0.1334878701890412, "grad_norm": 0.18896661698818207, "learning_rate": 8.34875e-05, "loss": 0.0522, "step": 6680 }, { "epoch": 0.13368770233004276, "grad_norm": 0.16970638930797577, "learning_rate": 8.36125e-05, "loss": 0.0545, "step": 6690 }, { "epoch": 0.13388753447104432, "grad_norm": 0.18621447682380676, "learning_rate": 8.37375e-05, "loss": 0.0618, "step": 6700 }, { "epoch": 0.13408736661204587, "grad_norm": 0.21419210731983185, "learning_rate": 8.38625e-05, "loss": 0.0542, "step": 6710 }, { "epoch": 0.13428719875304743, "grad_norm": 0.2218618094921112, "learning_rate": 8.39875e-05, "loss": 0.0537, "step": 6720 }, { "epoch": 0.134487030894049, "grad_norm": 0.20652513206005096, "learning_rate": 8.41125e-05, "loss": 0.066, "step": 6730 }, { "epoch": 0.13468686303505056, "grad_norm": 0.2121414989233017, "learning_rate": 8.423750000000001e-05, "loss": 0.0619, "step": 6740 }, { "epoch": 0.13488669517605212, "grad_norm": 0.24746213853359222, "learning_rate": 8.43625e-05, "loss": 0.0516, "step": 6750 }, { "epoch": 0.13508652731705367, "grad_norm": 0.21296823024749756, "learning_rate": 8.448750000000001e-05, "loss": 0.0646, "step": 6760 }, { "epoch": 0.13528635945805523, "grad_norm": 0.2293241024017334, "learning_rate": 8.46125e-05, "loss": 0.0603, "step": 6770 }, { "epoch": 0.13548619159905678, "grad_norm": 0.2750142514705658, "learning_rate": 8.47375e-05, "loss": 0.0564, "step": 6780 }, { "epoch": 0.13568602374005836, "grad_norm": 0.157621830701828, "learning_rate": 8.48625e-05, "loss": 0.0514, "step": 6790 }, { "epoch": 0.13588585588105992, "grad_norm": 0.20974451303482056, "learning_rate": 8.498750000000001e-05, "loss": 0.0736, "step": 6800 }, { "epoch": 0.13608568802206147, "grad_norm": 0.22295598685741425, "learning_rate": 8.51125e-05, "loss": 0.0585, "step": 6810 }, { "epoch": 0.13628552016306303, "grad_norm": 0.2296207845211029, "learning_rate": 8.52375e-05, "loss": 0.0506, "step": 6820 }, { "epoch": 0.13648535230406458, "grad_norm": 0.18972377479076385, "learning_rate": 8.53625e-05, "loss": 0.0667, "step": 6830 }, { "epoch": 0.13668518444506614, "grad_norm": 0.15406498312950134, "learning_rate": 8.548750000000001e-05, "loss": 0.0514, "step": 6840 }, { "epoch": 0.1368850165860677, "grad_norm": 0.22539211809635162, "learning_rate": 8.56125e-05, "loss": 0.0548, "step": 6850 }, { "epoch": 0.13708484872706928, "grad_norm": 0.17796482145786285, "learning_rate": 8.57375e-05, "loss": 0.0497, "step": 6860 }, { "epoch": 0.13728468086807083, "grad_norm": 0.13701798021793365, "learning_rate": 8.58625e-05, "loss": 0.0473, "step": 6870 }, { "epoch": 0.13748451300907238, "grad_norm": 0.18007700145244598, "learning_rate": 8.59875e-05, "loss": 0.0633, "step": 6880 }, { "epoch": 0.13768434515007394, "grad_norm": 0.21236121654510498, "learning_rate": 8.611250000000001e-05, "loss": 0.0512, "step": 6890 }, { "epoch": 0.1378841772910755, "grad_norm": 0.17369425296783447, "learning_rate": 8.62375e-05, "loss": 0.0526, "step": 6900 }, { "epoch": 0.13808400943207705, "grad_norm": 0.17906805872917175, "learning_rate": 8.63625e-05, "loss": 0.0571, "step": 6910 }, { "epoch": 0.1382838415730786, "grad_norm": 0.22749286890029907, "learning_rate": 8.64875e-05, "loss": 0.0616, "step": 6920 }, { "epoch": 0.13848367371408019, "grad_norm": 0.1563132107257843, "learning_rate": 8.661250000000001e-05, "loss": 0.0517, "step": 6930 }, { "epoch": 0.13868350585508174, "grad_norm": 0.15081501007080078, "learning_rate": 8.67375e-05, "loss": 0.0507, "step": 6940 }, { "epoch": 0.1388833379960833, "grad_norm": 0.1770065426826477, "learning_rate": 8.68625e-05, "loss": 0.0554, "step": 6950 }, { "epoch": 0.13908317013708485, "grad_norm": 0.20951798558235168, "learning_rate": 8.69875e-05, "loss": 0.0598, "step": 6960 }, { "epoch": 0.1392830022780864, "grad_norm": 0.19773785769939423, "learning_rate": 8.711250000000001e-05, "loss": 0.0569, "step": 6970 }, { "epoch": 0.13948283441908796, "grad_norm": 0.18821851909160614, "learning_rate": 8.72375e-05, "loss": 0.055, "step": 6980 }, { "epoch": 0.1396826665600895, "grad_norm": 0.19006215035915375, "learning_rate": 8.73625e-05, "loss": 0.0698, "step": 6990 }, { "epoch": 0.1398824987010911, "grad_norm": 0.22523760795593262, "learning_rate": 8.74875e-05, "loss": 0.0603, "step": 7000 }, { "epoch": 0.14008233084209265, "grad_norm": 0.2716164290904999, "learning_rate": 8.761250000000001e-05, "loss": 0.0619, "step": 7010 }, { "epoch": 0.1402821629830942, "grad_norm": 0.17825406789779663, "learning_rate": 8.773750000000001e-05, "loss": 0.0562, "step": 7020 }, { "epoch": 0.14048199512409576, "grad_norm": 0.17992615699768066, "learning_rate": 8.78625e-05, "loss": 0.0507, "step": 7030 }, { "epoch": 0.14068182726509731, "grad_norm": 0.20264123380184174, "learning_rate": 8.79875e-05, "loss": 0.0523, "step": 7040 }, { "epoch": 0.14088165940609887, "grad_norm": 0.21389231085777283, "learning_rate": 8.811250000000001e-05, "loss": 0.0524, "step": 7050 }, { "epoch": 0.14108149154710042, "grad_norm": 0.1729029417037964, "learning_rate": 8.823750000000001e-05, "loss": 0.0547, "step": 7060 }, { "epoch": 0.141281323688102, "grad_norm": 0.1886783242225647, "learning_rate": 8.83625e-05, "loss": 0.0589, "step": 7070 }, { "epoch": 0.14148115582910356, "grad_norm": 0.24211828410625458, "learning_rate": 8.84875e-05, "loss": 0.0606, "step": 7080 }, { "epoch": 0.14168098797010512, "grad_norm": 0.20397023856639862, "learning_rate": 8.86125e-05, "loss": 0.0562, "step": 7090 }, { "epoch": 0.14188082011110667, "grad_norm": 0.16103309392929077, "learning_rate": 8.873750000000001e-05, "loss": 0.0592, "step": 7100 }, { "epoch": 0.14208065225210822, "grad_norm": 0.18412962555885315, "learning_rate": 8.88625e-05, "loss": 0.0583, "step": 7110 }, { "epoch": 0.14228048439310978, "grad_norm": 0.19700583815574646, "learning_rate": 8.89875e-05, "loss": 0.0544, "step": 7120 }, { "epoch": 0.14248031653411133, "grad_norm": 0.196736678481102, "learning_rate": 8.91125e-05, "loss": 0.0592, "step": 7130 }, { "epoch": 0.14268014867511292, "grad_norm": 0.22435501217842102, "learning_rate": 8.923750000000001e-05, "loss": 0.0654, "step": 7140 }, { "epoch": 0.14287998081611447, "grad_norm": 0.20021763443946838, "learning_rate": 8.936250000000001e-05, "loss": 0.0547, "step": 7150 }, { "epoch": 0.14307981295711603, "grad_norm": 0.22144649922847748, "learning_rate": 8.94875e-05, "loss": 0.0604, "step": 7160 }, { "epoch": 0.14327964509811758, "grad_norm": 0.22556546330451965, "learning_rate": 8.96125e-05, "loss": 0.0549, "step": 7170 }, { "epoch": 0.14347947723911914, "grad_norm": 0.20483984053134918, "learning_rate": 8.97375e-05, "loss": 0.0517, "step": 7180 }, { "epoch": 0.1436793093801207, "grad_norm": 0.15584199130535126, "learning_rate": 8.986250000000001e-05, "loss": 0.0572, "step": 7190 }, { "epoch": 0.14387914152112224, "grad_norm": 0.16075554490089417, "learning_rate": 8.99875e-05, "loss": 0.0599, "step": 7200 }, { "epoch": 0.14407897366212383, "grad_norm": 0.19278520345687866, "learning_rate": 9.01125e-05, "loss": 0.0537, "step": 7210 }, { "epoch": 0.14427880580312538, "grad_norm": 0.21524880826473236, "learning_rate": 9.02375e-05, "loss": 0.0529, "step": 7220 }, { "epoch": 0.14447863794412694, "grad_norm": 0.18170641362667084, "learning_rate": 9.036250000000001e-05, "loss": 0.056, "step": 7230 }, { "epoch": 0.1446784700851285, "grad_norm": 0.16156482696533203, "learning_rate": 9.04875e-05, "loss": 0.0552, "step": 7240 }, { "epoch": 0.14487830222613005, "grad_norm": 0.20758958160877228, "learning_rate": 9.06125e-05, "loss": 0.0601, "step": 7250 }, { "epoch": 0.1450781343671316, "grad_norm": 0.1575046181678772, "learning_rate": 9.07375e-05, "loss": 0.0539, "step": 7260 }, { "epoch": 0.14527796650813316, "grad_norm": 0.18334738910198212, "learning_rate": 9.08625e-05, "loss": 0.062, "step": 7270 }, { "epoch": 0.14547779864913474, "grad_norm": 0.21118281781673431, "learning_rate": 9.098750000000001e-05, "loss": 0.051, "step": 7280 }, { "epoch": 0.1456776307901363, "grad_norm": 0.2179885357618332, "learning_rate": 9.11125e-05, "loss": 0.0519, "step": 7290 }, { "epoch": 0.14587746293113785, "grad_norm": 0.24898158013820648, "learning_rate": 9.123750000000002e-05, "loss": 0.0583, "step": 7300 }, { "epoch": 0.1460772950721394, "grad_norm": 0.17766252160072327, "learning_rate": 9.13625e-05, "loss": 0.0493, "step": 7310 }, { "epoch": 0.14627712721314096, "grad_norm": 0.2660791277885437, "learning_rate": 9.148750000000001e-05, "loss": 0.0506, "step": 7320 }, { "epoch": 0.1464769593541425, "grad_norm": 0.21121278405189514, "learning_rate": 9.16125e-05, "loss": 0.0538, "step": 7330 }, { "epoch": 0.14667679149514407, "grad_norm": 0.19827841222286224, "learning_rate": 9.173750000000001e-05, "loss": 0.0589, "step": 7340 }, { "epoch": 0.14687662363614565, "grad_norm": 0.212886244058609, "learning_rate": 9.18625e-05, "loss": 0.0593, "step": 7350 }, { "epoch": 0.1470764557771472, "grad_norm": 0.22346854209899902, "learning_rate": 9.19875e-05, "loss": 0.0628, "step": 7360 }, { "epoch": 0.14727628791814876, "grad_norm": 0.1495327353477478, "learning_rate": 9.21125e-05, "loss": 0.0612, "step": 7370 }, { "epoch": 0.1474761200591503, "grad_norm": 0.17944008111953735, "learning_rate": 9.22375e-05, "loss": 0.061, "step": 7380 }, { "epoch": 0.14767595220015187, "grad_norm": 0.18711315095424652, "learning_rate": 9.23625e-05, "loss": 0.0542, "step": 7390 }, { "epoch": 0.14787578434115342, "grad_norm": 0.21746207773685455, "learning_rate": 9.24875e-05, "loss": 0.0533, "step": 7400 }, { "epoch": 0.148075616482155, "grad_norm": 0.2274678647518158, "learning_rate": 9.261250000000001e-05, "loss": 0.0527, "step": 7410 }, { "epoch": 0.14827544862315656, "grad_norm": 0.16415086388587952, "learning_rate": 9.27375e-05, "loss": 0.0633, "step": 7420 }, { "epoch": 0.1484752807641581, "grad_norm": 0.28988009691238403, "learning_rate": 9.286250000000001e-05, "loss": 0.0568, "step": 7430 }, { "epoch": 0.14867511290515967, "grad_norm": 0.19010628759860992, "learning_rate": 9.29875e-05, "loss": 0.0552, "step": 7440 }, { "epoch": 0.14887494504616122, "grad_norm": 0.17375226318836212, "learning_rate": 9.311250000000001e-05, "loss": 0.0507, "step": 7450 }, { "epoch": 0.14907477718716278, "grad_norm": 0.17793779075145721, "learning_rate": 9.32375e-05, "loss": 0.0529, "step": 7460 }, { "epoch": 0.14927460932816433, "grad_norm": 0.2165568619966507, "learning_rate": 9.336250000000001e-05, "loss": 0.0573, "step": 7470 }, { "epoch": 0.14947444146916591, "grad_norm": 0.18271830677986145, "learning_rate": 9.34875e-05, "loss": 0.0531, "step": 7480 }, { "epoch": 0.14967427361016747, "grad_norm": 0.1274200677871704, "learning_rate": 9.36125e-05, "loss": 0.0541, "step": 7490 }, { "epoch": 0.14987410575116902, "grad_norm": 0.1852603405714035, "learning_rate": 9.37375e-05, "loss": 0.0581, "step": 7500 }, { "epoch": 0.15007393789217058, "grad_norm": 0.26552778482437134, "learning_rate": 9.386250000000001e-05, "loss": 0.0483, "step": 7510 }, { "epoch": 0.15027377003317213, "grad_norm": 0.22061476111412048, "learning_rate": 9.39875e-05, "loss": 0.0604, "step": 7520 }, { "epoch": 0.1504736021741737, "grad_norm": 0.16039520502090454, "learning_rate": 9.41125e-05, "loss": 0.0667, "step": 7530 }, { "epoch": 0.15067343431517524, "grad_norm": 0.1532737761735916, "learning_rate": 9.423750000000001e-05, "loss": 0.0581, "step": 7540 }, { "epoch": 0.15087326645617682, "grad_norm": 0.27276596426963806, "learning_rate": 9.436250000000001e-05, "loss": 0.0609, "step": 7550 }, { "epoch": 0.15107309859717838, "grad_norm": 0.17862097918987274, "learning_rate": 9.448750000000001e-05, "loss": 0.0523, "step": 7560 }, { "epoch": 0.15127293073817993, "grad_norm": 0.19563686847686768, "learning_rate": 9.46125e-05, "loss": 0.0468, "step": 7570 }, { "epoch": 0.1514727628791815, "grad_norm": 0.18599353730678558, "learning_rate": 9.47375e-05, "loss": 0.0591, "step": 7580 }, { "epoch": 0.15167259502018304, "grad_norm": 0.13201561570167542, "learning_rate": 9.486250000000001e-05, "loss": 0.055, "step": 7590 }, { "epoch": 0.1518724271611846, "grad_norm": 0.20798569917678833, "learning_rate": 9.498750000000001e-05, "loss": 0.0574, "step": 7600 }, { "epoch": 0.15207225930218615, "grad_norm": 0.17234665155410767, "learning_rate": 9.51125e-05, "loss": 0.0541, "step": 7610 }, { "epoch": 0.15227209144318773, "grad_norm": 0.1880452036857605, "learning_rate": 9.52375e-05, "loss": 0.049, "step": 7620 }, { "epoch": 0.1524719235841893, "grad_norm": 0.19480830430984497, "learning_rate": 9.53625e-05, "loss": 0.0681, "step": 7630 }, { "epoch": 0.15267175572519084, "grad_norm": 0.1824312061071396, "learning_rate": 9.548750000000001e-05, "loss": 0.0505, "step": 7640 }, { "epoch": 0.1528715878661924, "grad_norm": 0.2138872742652893, "learning_rate": 9.56125e-05, "loss": 0.0642, "step": 7650 }, { "epoch": 0.15307142000719395, "grad_norm": 0.15839433670043945, "learning_rate": 9.57375e-05, "loss": 0.053, "step": 7660 }, { "epoch": 0.1532712521481955, "grad_norm": 0.1789252907037735, "learning_rate": 9.58625e-05, "loss": 0.0551, "step": 7670 }, { "epoch": 0.15347108428919706, "grad_norm": 0.13174763321876526, "learning_rate": 9.598750000000001e-05, "loss": 0.055, "step": 7680 }, { "epoch": 0.15367091643019865, "grad_norm": 0.14308291673660278, "learning_rate": 9.611250000000001e-05, "loss": 0.0526, "step": 7690 }, { "epoch": 0.1538707485712002, "grad_norm": 0.14122791588306427, "learning_rate": 9.62375e-05, "loss": 0.0512, "step": 7700 }, { "epoch": 0.15407058071220175, "grad_norm": 0.17569544911384583, "learning_rate": 9.63625e-05, "loss": 0.0533, "step": 7710 }, { "epoch": 0.1542704128532033, "grad_norm": 0.180332213640213, "learning_rate": 9.648750000000001e-05, "loss": 0.064, "step": 7720 }, { "epoch": 0.15447024499420486, "grad_norm": 0.16203732788562775, "learning_rate": 9.661250000000001e-05, "loss": 0.0601, "step": 7730 }, { "epoch": 0.15467007713520642, "grad_norm": 0.27122417092323303, "learning_rate": 9.67375e-05, "loss": 0.0559, "step": 7740 }, { "epoch": 0.15486990927620797, "grad_norm": 0.19553150236606598, "learning_rate": 9.68625e-05, "loss": 0.0591, "step": 7750 }, { "epoch": 0.15506974141720956, "grad_norm": 0.1489051878452301, "learning_rate": 9.69875e-05, "loss": 0.0514, "step": 7760 }, { "epoch": 0.1552695735582111, "grad_norm": 0.2375158667564392, "learning_rate": 9.711250000000001e-05, "loss": 0.0489, "step": 7770 }, { "epoch": 0.15546940569921266, "grad_norm": 0.2573506236076355, "learning_rate": 9.72375e-05, "loss": 0.0561, "step": 7780 }, { "epoch": 0.15566923784021422, "grad_norm": 0.17351044714450836, "learning_rate": 9.73625e-05, "loss": 0.066, "step": 7790 }, { "epoch": 0.15586906998121577, "grad_norm": 0.21229307353496552, "learning_rate": 9.74875e-05, "loss": 0.0495, "step": 7800 }, { "epoch": 0.15606890212221733, "grad_norm": 0.15807116031646729, "learning_rate": 9.761250000000001e-05, "loss": 0.0565, "step": 7810 }, { "epoch": 0.15626873426321888, "grad_norm": 0.20450431108474731, "learning_rate": 9.773750000000001e-05, "loss": 0.0572, "step": 7820 }, { "epoch": 0.15646856640422047, "grad_norm": 0.17882578074932098, "learning_rate": 9.78625e-05, "loss": 0.0588, "step": 7830 }, { "epoch": 0.15666839854522202, "grad_norm": 0.16596727073192596, "learning_rate": 9.79875e-05, "loss": 0.0476, "step": 7840 }, { "epoch": 0.15686823068622358, "grad_norm": 0.16115860641002655, "learning_rate": 9.81125e-05, "loss": 0.0555, "step": 7850 }, { "epoch": 0.15706806282722513, "grad_norm": 0.1369127333164215, "learning_rate": 9.823750000000001e-05, "loss": 0.0557, "step": 7860 }, { "epoch": 0.15726789496822668, "grad_norm": 0.1862086057662964, "learning_rate": 9.83625e-05, "loss": 0.0605, "step": 7870 }, { "epoch": 0.15746772710922824, "grad_norm": 0.17642593383789062, "learning_rate": 9.84875e-05, "loss": 0.0574, "step": 7880 }, { "epoch": 0.1576675592502298, "grad_norm": 0.15751424431800842, "learning_rate": 9.86125e-05, "loss": 0.0559, "step": 7890 }, { "epoch": 0.15786739139123138, "grad_norm": 0.2103520780801773, "learning_rate": 9.873750000000001e-05, "loss": 0.0636, "step": 7900 }, { "epoch": 0.15806722353223293, "grad_norm": 0.14988602697849274, "learning_rate": 9.88625e-05, "loss": 0.0541, "step": 7910 }, { "epoch": 0.15826705567323449, "grad_norm": 0.1505022644996643, "learning_rate": 9.89875e-05, "loss": 0.0529, "step": 7920 }, { "epoch": 0.15846688781423604, "grad_norm": 0.19196195900440216, "learning_rate": 9.91125e-05, "loss": 0.0561, "step": 7930 }, { "epoch": 0.1586667199552376, "grad_norm": 0.18247130513191223, "learning_rate": 9.92375e-05, "loss": 0.0637, "step": 7940 }, { "epoch": 0.15886655209623915, "grad_norm": 0.1670960634946823, "learning_rate": 9.93625e-05, "loss": 0.0521, "step": 7950 }, { "epoch": 0.1590663842372407, "grad_norm": 0.1651657521724701, "learning_rate": 9.94875e-05, "loss": 0.0561, "step": 7960 }, { "epoch": 0.1592662163782423, "grad_norm": 0.1953902542591095, "learning_rate": 9.96125e-05, "loss": 0.0565, "step": 7970 }, { "epoch": 0.15946604851924384, "grad_norm": 0.2596529424190521, "learning_rate": 9.97375e-05, "loss": 0.0513, "step": 7980 }, { "epoch": 0.1596658806602454, "grad_norm": 0.18464459478855133, "learning_rate": 9.986250000000001e-05, "loss": 0.058, "step": 7990 }, { "epoch": 0.15986571280124695, "grad_norm": 0.176949143409729, "learning_rate": 9.99875e-05, "loss": 0.0603, "step": 8000 }, { "epoch": 0.1600655449422485, "grad_norm": 0.15856388211250305, "learning_rate": 9.999999913495721e-05, "loss": 0.0517, "step": 8010 }, { "epoch": 0.16026537708325006, "grad_norm": 0.19402873516082764, "learning_rate": 9.999999614468584e-05, "loss": 0.0544, "step": 8020 }, { "epoch": 0.16046520922425161, "grad_norm": 0.17089511454105377, "learning_rate": 9.999999101850648e-05, "loss": 0.0524, "step": 8030 }, { "epoch": 0.1606650413652532, "grad_norm": 0.20497727394104004, "learning_rate": 9.999998375641937e-05, "loss": 0.0552, "step": 8040 }, { "epoch": 0.16086487350625475, "grad_norm": 0.2304200977087021, "learning_rate": 9.99999743584248e-05, "loss": 0.0604, "step": 8050 }, { "epoch": 0.1610647056472563, "grad_norm": 0.1989251971244812, "learning_rate": 9.999996282452317e-05, "loss": 0.0612, "step": 8060 }, { "epoch": 0.16126453778825786, "grad_norm": 0.20095610618591309, "learning_rate": 9.9999949154715e-05, "loss": 0.0562, "step": 8070 }, { "epoch": 0.16146436992925942, "grad_norm": 0.24551309645175934, "learning_rate": 9.999993334900085e-05, "loss": 0.0602, "step": 8080 }, { "epoch": 0.16166420207026097, "grad_norm": 0.1684368997812271, "learning_rate": 9.99999154073814e-05, "loss": 0.0479, "step": 8090 }, { "epoch": 0.16186403421126255, "grad_norm": 0.1696418672800064, "learning_rate": 9.99998953298574e-05, "loss": 0.0551, "step": 8100 }, { "epoch": 0.1620638663522641, "grad_norm": 0.19372659921646118, "learning_rate": 9.999987311642974e-05, "loss": 0.0545, "step": 8110 }, { "epoch": 0.16226369849326566, "grad_norm": 0.21617916226387024, "learning_rate": 9.999984876709936e-05, "loss": 0.0671, "step": 8120 }, { "epoch": 0.16246353063426722, "grad_norm": 0.2539534270763397, "learning_rate": 9.999982228186727e-05, "loss": 0.0603, "step": 8130 }, { "epoch": 0.16266336277526877, "grad_norm": 0.19184599816799164, "learning_rate": 9.999979366073464e-05, "loss": 0.0534, "step": 8140 }, { "epoch": 0.16286319491627033, "grad_norm": 0.22170332074165344, "learning_rate": 9.999976290370267e-05, "loss": 0.0511, "step": 8150 }, { "epoch": 0.16306302705727188, "grad_norm": 0.21905161440372467, "learning_rate": 9.99997300107727e-05, "loss": 0.0563, "step": 8160 }, { "epoch": 0.16326285919827346, "grad_norm": 0.2188177853822708, "learning_rate": 9.99996949819461e-05, "loss": 0.0566, "step": 8170 }, { "epoch": 0.16346269133927502, "grad_norm": 0.1510675549507141, "learning_rate": 9.999965781722437e-05, "loss": 0.0512, "step": 8180 }, { "epoch": 0.16366252348027657, "grad_norm": 0.17282339930534363, "learning_rate": 9.999961851660913e-05, "loss": 0.0591, "step": 8190 }, { "epoch": 0.16386235562127813, "grad_norm": 0.19559626281261444, "learning_rate": 9.999957708010205e-05, "loss": 0.0609, "step": 8200 }, { "epoch": 0.16406218776227968, "grad_norm": 0.20621083676815033, "learning_rate": 9.999953350770487e-05, "loss": 0.0575, "step": 8210 }, { "epoch": 0.16426201990328124, "grad_norm": 0.19912143051624298, "learning_rate": 9.999948779941947e-05, "loss": 0.0593, "step": 8220 }, { "epoch": 0.1644618520442828, "grad_norm": 0.21422940492630005, "learning_rate": 9.999943995524781e-05, "loss": 0.0517, "step": 8230 }, { "epoch": 0.16466168418528437, "grad_norm": 0.1784706562757492, "learning_rate": 9.999938997519194e-05, "loss": 0.0568, "step": 8240 }, { "epoch": 0.16486151632628593, "grad_norm": 0.18492355942726135, "learning_rate": 9.999933785925395e-05, "loss": 0.0576, "step": 8250 }, { "epoch": 0.16506134846728748, "grad_norm": 0.2252187877893448, "learning_rate": 9.999928360743612e-05, "loss": 0.058, "step": 8260 }, { "epoch": 0.16526118060828904, "grad_norm": 0.14514954388141632, "learning_rate": 9.999922721974075e-05, "loss": 0.0541, "step": 8270 }, { "epoch": 0.1654610127492906, "grad_norm": 0.1777399182319641, "learning_rate": 9.999916869617024e-05, "loss": 0.0635, "step": 8280 }, { "epoch": 0.16566084489029215, "grad_norm": 0.23109561204910278, "learning_rate": 9.999910803672708e-05, "loss": 0.061, "step": 8290 }, { "epoch": 0.1658606770312937, "grad_norm": 0.17603181302547455, "learning_rate": 9.999904524141389e-05, "loss": 0.0464, "step": 8300 }, { "epoch": 0.16606050917229528, "grad_norm": 0.22792260348796844, "learning_rate": 9.999898031023334e-05, "loss": 0.0448, "step": 8310 }, { "epoch": 0.16626034131329684, "grad_norm": 0.15529893338680267, "learning_rate": 9.99989132431882e-05, "loss": 0.0524, "step": 8320 }, { "epoch": 0.1664601734542984, "grad_norm": 0.15251661837100983, "learning_rate": 9.999884404028133e-05, "loss": 0.0579, "step": 8330 }, { "epoch": 0.16666000559529995, "grad_norm": 0.2081565111875534, "learning_rate": 9.99987727015157e-05, "loss": 0.0575, "step": 8340 }, { "epoch": 0.1668598377363015, "grad_norm": 0.1559692919254303, "learning_rate": 9.999869922689435e-05, "loss": 0.055, "step": 8350 }, { "epoch": 0.16705966987730306, "grad_norm": 0.130392387509346, "learning_rate": 9.999862361642041e-05, "loss": 0.0646, "step": 8360 }, { "epoch": 0.1672595020183046, "grad_norm": 0.14700278639793396, "learning_rate": 9.999854587009714e-05, "loss": 0.0558, "step": 8370 }, { "epoch": 0.1674593341593062, "grad_norm": 0.13470767438411713, "learning_rate": 9.999846598792782e-05, "loss": 0.0541, "step": 8380 }, { "epoch": 0.16765916630030775, "grad_norm": 0.258227676153183, "learning_rate": 9.99983839699159e-05, "loss": 0.0542, "step": 8390 }, { "epoch": 0.1678589984413093, "grad_norm": 0.29404622316360474, "learning_rate": 9.999829981606485e-05, "loss": 0.0515, "step": 8400 }, { "epoch": 0.16805883058231086, "grad_norm": 0.24309368431568146, "learning_rate": 9.999821352637829e-05, "loss": 0.0499, "step": 8410 }, { "epoch": 0.1682586627233124, "grad_norm": 0.19910019636154175, "learning_rate": 9.99981251008599e-05, "loss": 0.0505, "step": 8420 }, { "epoch": 0.16845849486431397, "grad_norm": 0.187799334526062, "learning_rate": 9.999803453951346e-05, "loss": 0.0566, "step": 8430 }, { "epoch": 0.16865832700531552, "grad_norm": 0.23852813243865967, "learning_rate": 9.999794184234283e-05, "loss": 0.0493, "step": 8440 }, { "epoch": 0.1688581591463171, "grad_norm": 0.23115487396717072, "learning_rate": 9.999784700935196e-05, "loss": 0.0524, "step": 8450 }, { "epoch": 0.16905799128731866, "grad_norm": 0.11795562505722046, "learning_rate": 9.999775004054493e-05, "loss": 0.0492, "step": 8460 }, { "epoch": 0.16925782342832021, "grad_norm": 0.17387980222702026, "learning_rate": 9.999765093592586e-05, "loss": 0.0551, "step": 8470 }, { "epoch": 0.16945765556932177, "grad_norm": 0.5311224460601807, "learning_rate": 9.9997549695499e-05, "loss": 0.062, "step": 8480 }, { "epoch": 0.16965748771032332, "grad_norm": 0.12215887010097504, "learning_rate": 9.999744631926866e-05, "loss": 0.0577, "step": 8490 }, { "epoch": 0.16985731985132488, "grad_norm": 0.2107054889202118, "learning_rate": 9.999734080723925e-05, "loss": 0.0636, "step": 8500 }, { "epoch": 0.17005715199232643, "grad_norm": 0.17408207058906555, "learning_rate": 9.99972331594153e-05, "loss": 0.0647, "step": 8510 }, { "epoch": 0.17025698413332802, "grad_norm": 0.19186340272426605, "learning_rate": 9.999712337580141e-05, "loss": 0.0468, "step": 8520 }, { "epoch": 0.17045681627432957, "grad_norm": 0.1701149195432663, "learning_rate": 9.999701145640223e-05, "loss": 0.0552, "step": 8530 }, { "epoch": 0.17065664841533112, "grad_norm": 0.17869937419891357, "learning_rate": 9.999689740122258e-05, "loss": 0.0479, "step": 8540 }, { "epoch": 0.17085648055633268, "grad_norm": 0.1437879055738449, "learning_rate": 9.999678121026733e-05, "loss": 0.0529, "step": 8550 }, { "epoch": 0.17105631269733423, "grad_norm": 0.1673593521118164, "learning_rate": 9.999666288354143e-05, "loss": 0.0553, "step": 8560 }, { "epoch": 0.1712561448383358, "grad_norm": 0.1618250161409378, "learning_rate": 9.999654242104993e-05, "loss": 0.0466, "step": 8570 }, { "epoch": 0.17145597697933734, "grad_norm": 0.1869051307439804, "learning_rate": 9.9996419822798e-05, "loss": 0.0588, "step": 8580 }, { "epoch": 0.17165580912033893, "grad_norm": 0.15475600957870483, "learning_rate": 9.999629508879084e-05, "loss": 0.0551, "step": 8590 }, { "epoch": 0.17185564126134048, "grad_norm": 0.18265755474567413, "learning_rate": 9.999616821903382e-05, "loss": 0.0516, "step": 8600 }, { "epoch": 0.17205547340234204, "grad_norm": 0.1436328887939453, "learning_rate": 9.99960392135323e-05, "loss": 0.0504, "step": 8610 }, { "epoch": 0.1722553055433436, "grad_norm": 0.19080224633216858, "learning_rate": 9.999590807229187e-05, "loss": 0.0506, "step": 8620 }, { "epoch": 0.17245513768434514, "grad_norm": 0.21762071549892426, "learning_rate": 9.99957747953181e-05, "loss": 0.0571, "step": 8630 }, { "epoch": 0.1726549698253467, "grad_norm": 0.17128875851631165, "learning_rate": 9.999563938261664e-05, "loss": 0.0465, "step": 8640 }, { "epoch": 0.17285480196634825, "grad_norm": 0.15804548561573029, "learning_rate": 9.999550183419333e-05, "loss": 0.0549, "step": 8650 }, { "epoch": 0.17305463410734984, "grad_norm": 0.19678443670272827, "learning_rate": 9.999536215005403e-05, "loss": 0.0473, "step": 8660 }, { "epoch": 0.1732544662483514, "grad_norm": 0.14784817397594452, "learning_rate": 9.99952203302047e-05, "loss": 0.0522, "step": 8670 }, { "epoch": 0.17345429838935295, "grad_norm": 0.16228564083576202, "learning_rate": 9.99950763746514e-05, "loss": 0.0545, "step": 8680 }, { "epoch": 0.1736541305303545, "grad_norm": 0.18951751291751862, "learning_rate": 9.999493028340027e-05, "loss": 0.0529, "step": 8690 }, { "epoch": 0.17385396267135605, "grad_norm": 0.18327696621418, "learning_rate": 9.999478205645758e-05, "loss": 0.0545, "step": 8700 }, { "epoch": 0.1740537948123576, "grad_norm": 0.1913805902004242, "learning_rate": 9.999463169382964e-05, "loss": 0.0492, "step": 8710 }, { "epoch": 0.1742536269533592, "grad_norm": 0.1693105250597, "learning_rate": 9.999447919552287e-05, "loss": 0.0525, "step": 8720 }, { "epoch": 0.17445345909436075, "grad_norm": 0.14329800009727478, "learning_rate": 9.999432456154381e-05, "loss": 0.0548, "step": 8730 }, { "epoch": 0.1746532912353623, "grad_norm": 0.194334015250206, "learning_rate": 9.999416779189903e-05, "loss": 0.0518, "step": 8740 }, { "epoch": 0.17485312337636386, "grad_norm": 0.18406054377555847, "learning_rate": 9.999400888659524e-05, "loss": 0.0467, "step": 8750 }, { "epoch": 0.1750529555173654, "grad_norm": 0.33278778195381165, "learning_rate": 9.999384784563924e-05, "loss": 0.0566, "step": 8760 }, { "epoch": 0.17525278765836697, "grad_norm": 0.15118952095508575, "learning_rate": 9.99936846690379e-05, "loss": 0.0599, "step": 8770 }, { "epoch": 0.17545261979936852, "grad_norm": 0.16761647164821625, "learning_rate": 9.99935193567982e-05, "loss": 0.0524, "step": 8780 }, { "epoch": 0.1756524519403701, "grad_norm": 0.23716041445732117, "learning_rate": 9.999335190892719e-05, "loss": 0.0642, "step": 8790 }, { "epoch": 0.17585228408137166, "grad_norm": 0.22060240805149078, "learning_rate": 9.999318232543202e-05, "loss": 0.0678, "step": 8800 }, { "epoch": 0.1760521162223732, "grad_norm": 0.19904851913452148, "learning_rate": 9.999301060631994e-05, "loss": 0.0586, "step": 8810 }, { "epoch": 0.17625194836337477, "grad_norm": 0.1680493950843811, "learning_rate": 9.99928367515983e-05, "loss": 0.0522, "step": 8820 }, { "epoch": 0.17645178050437632, "grad_norm": 0.19345612823963165, "learning_rate": 9.99926607612745e-05, "loss": 0.0548, "step": 8830 }, { "epoch": 0.17665161264537788, "grad_norm": 0.1738532930612564, "learning_rate": 9.999248263535608e-05, "loss": 0.0566, "step": 8840 }, { "epoch": 0.17685144478637943, "grad_norm": 0.14119243621826172, "learning_rate": 9.999230237385065e-05, "loss": 0.0506, "step": 8850 }, { "epoch": 0.177051276927381, "grad_norm": 0.1635897010564804, "learning_rate": 9.999211997676588e-05, "loss": 0.0576, "step": 8860 }, { "epoch": 0.17725110906838257, "grad_norm": 0.17899982631206512, "learning_rate": 9.99919354441096e-05, "loss": 0.056, "step": 8870 }, { "epoch": 0.17745094120938412, "grad_norm": 0.16995981335639954, "learning_rate": 9.999174877588967e-05, "loss": 0.0676, "step": 8880 }, { "epoch": 0.17765077335038568, "grad_norm": 0.18035243451595306, "learning_rate": 9.999155997211407e-05, "loss": 0.0551, "step": 8890 }, { "epoch": 0.17785060549138723, "grad_norm": 0.2137915939092636, "learning_rate": 9.999136903279087e-05, "loss": 0.041, "step": 8900 }, { "epoch": 0.17805043763238879, "grad_norm": 0.16095788776874542, "learning_rate": 9.999117595792821e-05, "loss": 0.06, "step": 8910 }, { "epoch": 0.17825026977339034, "grad_norm": 0.1281609833240509, "learning_rate": 9.999098074753436e-05, "loss": 0.0511, "step": 8920 }, { "epoch": 0.17845010191439192, "grad_norm": 0.18948817253112793, "learning_rate": 9.999078340161765e-05, "loss": 0.054, "step": 8930 }, { "epoch": 0.17864993405539348, "grad_norm": 0.20476846396923065, "learning_rate": 9.999058392018649e-05, "loss": 0.0563, "step": 8940 }, { "epoch": 0.17884976619639503, "grad_norm": 0.13953456282615662, "learning_rate": 9.999038230324944e-05, "loss": 0.0522, "step": 8950 }, { "epoch": 0.1790495983373966, "grad_norm": 0.19802801311016083, "learning_rate": 9.99901785508151e-05, "loss": 0.0551, "step": 8960 }, { "epoch": 0.17924943047839814, "grad_norm": 0.1618763506412506, "learning_rate": 9.998997266289215e-05, "loss": 0.0528, "step": 8970 }, { "epoch": 0.1794492626193997, "grad_norm": 0.19989286363124847, "learning_rate": 9.998976463948942e-05, "loss": 0.0562, "step": 8980 }, { "epoch": 0.17964909476040125, "grad_norm": 0.18535515666007996, "learning_rate": 9.998955448061575e-05, "loss": 0.0528, "step": 8990 }, { "epoch": 0.17984892690140283, "grad_norm": 0.1875256597995758, "learning_rate": 9.998934218628017e-05, "loss": 0.052, "step": 9000 }, { "epoch": 0.1800487590424044, "grad_norm": 0.15878431499004364, "learning_rate": 9.998912775649171e-05, "loss": 0.0617, "step": 9010 }, { "epoch": 0.18024859118340594, "grad_norm": 0.17419558763504028, "learning_rate": 9.998891119125958e-05, "loss": 0.0619, "step": 9020 }, { "epoch": 0.1804484233244075, "grad_norm": 0.1872832179069519, "learning_rate": 9.998869249059295e-05, "loss": 0.0556, "step": 9030 }, { "epoch": 0.18064825546540905, "grad_norm": 0.18656229972839355, "learning_rate": 9.998847165450124e-05, "loss": 0.0506, "step": 9040 }, { "epoch": 0.1808480876064106, "grad_norm": 0.16106079518795013, "learning_rate": 9.998824868299384e-05, "loss": 0.0601, "step": 9050 }, { "epoch": 0.18104791974741216, "grad_norm": 0.16238537430763245, "learning_rate": 9.998802357608028e-05, "loss": 0.0537, "step": 9060 }, { "epoch": 0.18124775188841374, "grad_norm": 0.17044948041439056, "learning_rate": 9.99877963337702e-05, "loss": 0.0507, "step": 9070 }, { "epoch": 0.1814475840294153, "grad_norm": 0.1510743498802185, "learning_rate": 9.998756695607329e-05, "loss": 0.0537, "step": 9080 }, { "epoch": 0.18164741617041685, "grad_norm": 0.21663601696491241, "learning_rate": 9.998733544299934e-05, "loss": 0.0521, "step": 9090 }, { "epoch": 0.1818472483114184, "grad_norm": 0.14611119031906128, "learning_rate": 9.998710179455826e-05, "loss": 0.0468, "step": 9100 }, { "epoch": 0.18204708045241996, "grad_norm": 0.17623546719551086, "learning_rate": 9.998686601076001e-05, "loss": 0.0502, "step": 9110 }, { "epoch": 0.18224691259342152, "grad_norm": 0.16048529744148254, "learning_rate": 9.998662809161468e-05, "loss": 0.0505, "step": 9120 }, { "epoch": 0.18244674473442307, "grad_norm": 0.185541570186615, "learning_rate": 9.998638803713242e-05, "loss": 0.0549, "step": 9130 }, { "epoch": 0.18264657687542465, "grad_norm": 0.1451421082019806, "learning_rate": 9.998614584732349e-05, "loss": 0.0526, "step": 9140 }, { "epoch": 0.1828464090164262, "grad_norm": 0.19857856631278992, "learning_rate": 9.998590152219824e-05, "loss": 0.05, "step": 9150 }, { "epoch": 0.18304624115742776, "grad_norm": 0.18822254240512848, "learning_rate": 9.998565506176712e-05, "loss": 0.0572, "step": 9160 }, { "epoch": 0.18324607329842932, "grad_norm": 0.21781499683856964, "learning_rate": 9.998540646604062e-05, "loss": 0.0562, "step": 9170 }, { "epoch": 0.18344590543943087, "grad_norm": 0.15254993736743927, "learning_rate": 9.998515573502939e-05, "loss": 0.0523, "step": 9180 }, { "epoch": 0.18364573758043243, "grad_norm": 0.16333156824111938, "learning_rate": 9.998490286874413e-05, "loss": 0.0506, "step": 9190 }, { "epoch": 0.18384556972143398, "grad_norm": 0.15403711795806885, "learning_rate": 9.998464786719565e-05, "loss": 0.0566, "step": 9200 }, { "epoch": 0.18404540186243556, "grad_norm": 0.14838837087154388, "learning_rate": 9.998439073039483e-05, "loss": 0.062, "step": 9210 }, { "epoch": 0.18424523400343712, "grad_norm": 0.131006121635437, "learning_rate": 9.998413145835266e-05, "loss": 0.0474, "step": 9220 }, { "epoch": 0.18444506614443867, "grad_norm": 0.259406715631485, "learning_rate": 9.998387005108024e-05, "loss": 0.055, "step": 9230 }, { "epoch": 0.18464489828544023, "grad_norm": 0.18383553624153137, "learning_rate": 9.99836065085887e-05, "loss": 0.0567, "step": 9240 }, { "epoch": 0.18484473042644178, "grad_norm": 0.15517616271972656, "learning_rate": 9.998334083088932e-05, "loss": 0.0529, "step": 9250 }, { "epoch": 0.18504456256744334, "grad_norm": 0.19291388988494873, "learning_rate": 9.998307301799342e-05, "loss": 0.0535, "step": 9260 }, { "epoch": 0.1852443947084449, "grad_norm": 0.20134511590003967, "learning_rate": 9.998280306991248e-05, "loss": 0.0499, "step": 9270 }, { "epoch": 0.18544422684944648, "grad_norm": 0.1818956881761551, "learning_rate": 9.998253098665802e-05, "loss": 0.0601, "step": 9280 }, { "epoch": 0.18564405899044803, "grad_norm": 0.20240458846092224, "learning_rate": 9.998225676824164e-05, "loss": 0.0512, "step": 9290 }, { "epoch": 0.18584389113144958, "grad_norm": 0.15687426924705505, "learning_rate": 9.998198041467508e-05, "loss": 0.0579, "step": 9300 }, { "epoch": 0.18604372327245114, "grad_norm": 0.16029681265354156, "learning_rate": 9.998170192597014e-05, "loss": 0.0553, "step": 9310 }, { "epoch": 0.1862435554134527, "grad_norm": 0.18384288251399994, "learning_rate": 9.99814213021387e-05, "loss": 0.0691, "step": 9320 }, { "epoch": 0.18644338755445425, "grad_norm": 0.19269275665283203, "learning_rate": 9.998113854319278e-05, "loss": 0.0533, "step": 9330 }, { "epoch": 0.18664321969545583, "grad_norm": 0.14388811588287354, "learning_rate": 9.998085364914443e-05, "loss": 0.0548, "step": 9340 }, { "epoch": 0.18684305183645739, "grad_norm": 0.18062429130077362, "learning_rate": 9.998056662000583e-05, "loss": 0.0469, "step": 9350 }, { "epoch": 0.18704288397745894, "grad_norm": 0.15888690948486328, "learning_rate": 9.998027745578923e-05, "loss": 0.0455, "step": 9360 }, { "epoch": 0.1872427161184605, "grad_norm": 0.1796942800283432, "learning_rate": 9.997998615650701e-05, "loss": 0.0639, "step": 9370 }, { "epoch": 0.18744254825946205, "grad_norm": 0.20498409867286682, "learning_rate": 9.997969272217159e-05, "loss": 0.0489, "step": 9380 }, { "epoch": 0.1876423804004636, "grad_norm": 0.18413540720939636, "learning_rate": 9.997939715279551e-05, "loss": 0.0632, "step": 9390 }, { "epoch": 0.18784221254146516, "grad_norm": 0.15770164132118225, "learning_rate": 9.997909944839142e-05, "loss": 0.05, "step": 9400 }, { "epoch": 0.18804204468246674, "grad_norm": 0.16175605356693268, "learning_rate": 9.997879960897198e-05, "loss": 0.0541, "step": 9410 }, { "epoch": 0.1882418768234683, "grad_norm": 0.18105867505073547, "learning_rate": 9.997849763455006e-05, "loss": 0.0523, "step": 9420 }, { "epoch": 0.18844170896446985, "grad_norm": 0.22995929419994354, "learning_rate": 9.997819352513852e-05, "loss": 0.0612, "step": 9430 }, { "epoch": 0.1886415411054714, "grad_norm": 0.23697121441364288, "learning_rate": 9.997788728075039e-05, "loss": 0.054, "step": 9440 }, { "epoch": 0.18884137324647296, "grad_norm": 0.20562711358070374, "learning_rate": 9.997757890139871e-05, "loss": 0.0592, "step": 9450 }, { "epoch": 0.18904120538747451, "grad_norm": 0.1865912675857544, "learning_rate": 9.997726838709667e-05, "loss": 0.0471, "step": 9460 }, { "epoch": 0.18924103752847607, "grad_norm": 0.16143816709518433, "learning_rate": 9.997695573785754e-05, "loss": 0.051, "step": 9470 }, { "epoch": 0.18944086966947765, "grad_norm": 0.14532510936260223, "learning_rate": 9.997664095369468e-05, "loss": 0.047, "step": 9480 }, { "epoch": 0.1896407018104792, "grad_norm": 0.177092045545578, "learning_rate": 9.997632403462152e-05, "loss": 0.056, "step": 9490 }, { "epoch": 0.18984053395148076, "grad_norm": 0.2080054134130478, "learning_rate": 9.99760049806516e-05, "loss": 0.0547, "step": 9500 }, { "epoch": 0.19004036609248232, "grad_norm": 0.17118100821971893, "learning_rate": 9.997568379179855e-05, "loss": 0.0552, "step": 9510 }, { "epoch": 0.19024019823348387, "grad_norm": 0.18922153115272522, "learning_rate": 9.997536046807611e-05, "loss": 0.052, "step": 9520 }, { "epoch": 0.19044003037448543, "grad_norm": 0.2024574726819992, "learning_rate": 9.997503500949807e-05, "loss": 0.0562, "step": 9530 }, { "epoch": 0.19063986251548698, "grad_norm": 0.16258074343204498, "learning_rate": 9.997470741607835e-05, "loss": 0.0541, "step": 9540 }, { "epoch": 0.19083969465648856, "grad_norm": 0.1710330694913864, "learning_rate": 9.997437768783094e-05, "loss": 0.0592, "step": 9550 }, { "epoch": 0.19103952679749012, "grad_norm": 0.19153645634651184, "learning_rate": 9.997404582476992e-05, "loss": 0.0505, "step": 9560 }, { "epoch": 0.19123935893849167, "grad_norm": 0.18072421848773956, "learning_rate": 9.997371182690947e-05, "loss": 0.0539, "step": 9570 }, { "epoch": 0.19143919107949323, "grad_norm": 0.3093312978744507, "learning_rate": 9.997337569426384e-05, "loss": 0.0461, "step": 9580 }, { "epoch": 0.19163902322049478, "grad_norm": 0.18976865708827972, "learning_rate": 9.99730374268474e-05, "loss": 0.0477, "step": 9590 }, { "epoch": 0.19183885536149634, "grad_norm": 0.1849459558725357, "learning_rate": 9.997269702467462e-05, "loss": 0.0597, "step": 9600 }, { "epoch": 0.1920386875024979, "grad_norm": 0.17313748598098755, "learning_rate": 9.997235448776004e-05, "loss": 0.0428, "step": 9610 }, { "epoch": 0.19223851964349947, "grad_norm": 0.17071783542633057, "learning_rate": 9.997200981611825e-05, "loss": 0.0568, "step": 9620 }, { "epoch": 0.19243835178450103, "grad_norm": 0.15164825320243835, "learning_rate": 9.997166300976402e-05, "loss": 0.0509, "step": 9630 }, { "epoch": 0.19263818392550258, "grad_norm": 0.14217887818813324, "learning_rate": 9.997131406871213e-05, "loss": 0.0464, "step": 9640 }, { "epoch": 0.19283801606650414, "grad_norm": 0.21114122867584229, "learning_rate": 9.997096299297751e-05, "loss": 0.059, "step": 9650 }, { "epoch": 0.1930378482075057, "grad_norm": 0.19385825097560883, "learning_rate": 9.997060978257515e-05, "loss": 0.0604, "step": 9660 }, { "epoch": 0.19323768034850725, "grad_norm": 0.1651909202337265, "learning_rate": 9.997025443752015e-05, "loss": 0.0709, "step": 9670 }, { "epoch": 0.1934375124895088, "grad_norm": 0.16078737378120422, "learning_rate": 9.996989695782768e-05, "loss": 0.05, "step": 9680 }, { "epoch": 0.19363734463051038, "grad_norm": 0.14347296953201294, "learning_rate": 9.996953734351299e-05, "loss": 0.0494, "step": 9690 }, { "epoch": 0.19383717677151194, "grad_norm": 0.14206118881702423, "learning_rate": 9.996917559459148e-05, "loss": 0.0544, "step": 9700 }, { "epoch": 0.1940370089125135, "grad_norm": 0.16044257581233978, "learning_rate": 9.996881171107856e-05, "loss": 0.0439, "step": 9710 }, { "epoch": 0.19423684105351505, "grad_norm": 0.17395052313804626, "learning_rate": 9.996844569298983e-05, "loss": 0.0551, "step": 9720 }, { "epoch": 0.1944366731945166, "grad_norm": 0.1509086787700653, "learning_rate": 9.996807754034088e-05, "loss": 0.0513, "step": 9730 }, { "epoch": 0.19463650533551816, "grad_norm": 0.15245415270328522, "learning_rate": 9.996770725314746e-05, "loss": 0.0529, "step": 9740 }, { "epoch": 0.1948363374765197, "grad_norm": 0.185859814286232, "learning_rate": 9.996733483142538e-05, "loss": 0.0559, "step": 9750 }, { "epoch": 0.1950361696175213, "grad_norm": 0.13881833851337433, "learning_rate": 9.996696027519054e-05, "loss": 0.0542, "step": 9760 }, { "epoch": 0.19523600175852285, "grad_norm": 0.15250499546527863, "learning_rate": 9.996658358445894e-05, "loss": 0.0467, "step": 9770 }, { "epoch": 0.1954358338995244, "grad_norm": 0.16726267337799072, "learning_rate": 9.996620475924669e-05, "loss": 0.0618, "step": 9780 }, { "epoch": 0.19563566604052596, "grad_norm": 0.16496634483337402, "learning_rate": 9.996582379956998e-05, "loss": 0.0527, "step": 9790 }, { "epoch": 0.1958354981815275, "grad_norm": 0.15074284374713898, "learning_rate": 9.996544070544505e-05, "loss": 0.0563, "step": 9800 }, { "epoch": 0.19603533032252907, "grad_norm": 0.24488598108291626, "learning_rate": 9.996505547688828e-05, "loss": 0.0521, "step": 9810 }, { "epoch": 0.19623516246353062, "grad_norm": 0.1615900844335556, "learning_rate": 9.996466811391614e-05, "loss": 0.0505, "step": 9820 }, { "epoch": 0.1964349946045322, "grad_norm": 0.23564286530017853, "learning_rate": 9.996427861654517e-05, "loss": 0.0488, "step": 9830 }, { "epoch": 0.19663482674553376, "grad_norm": 0.17521288990974426, "learning_rate": 9.996388698479199e-05, "loss": 0.0511, "step": 9840 }, { "epoch": 0.1968346588865353, "grad_norm": 0.21217170357704163, "learning_rate": 9.996349321867336e-05, "loss": 0.0568, "step": 9850 }, { "epoch": 0.19703449102753687, "grad_norm": 0.15685081481933594, "learning_rate": 9.996309731820606e-05, "loss": 0.0598, "step": 9860 }, { "epoch": 0.19723432316853842, "grad_norm": 0.17978915572166443, "learning_rate": 9.996269928340705e-05, "loss": 0.0491, "step": 9870 }, { "epoch": 0.19743415530953998, "grad_norm": 0.15453806519508362, "learning_rate": 9.99622991142933e-05, "loss": 0.0547, "step": 9880 }, { "epoch": 0.19763398745054153, "grad_norm": 0.17559054493904114, "learning_rate": 9.996189681088193e-05, "loss": 0.0535, "step": 9890 }, { "epoch": 0.19783381959154311, "grad_norm": 0.15754112601280212, "learning_rate": 9.99614923731901e-05, "loss": 0.0574, "step": 9900 }, { "epoch": 0.19803365173254467, "grad_norm": 0.19587254524230957, "learning_rate": 9.996108580123508e-05, "loss": 0.0546, "step": 9910 }, { "epoch": 0.19823348387354622, "grad_norm": 0.169503316283226, "learning_rate": 9.99606770950343e-05, "loss": 0.0443, "step": 9920 }, { "epoch": 0.19843331601454778, "grad_norm": 0.2075446993112564, "learning_rate": 9.996026625460514e-05, "loss": 0.0606, "step": 9930 }, { "epoch": 0.19863314815554933, "grad_norm": 0.16517114639282227, "learning_rate": 9.99598532799652e-05, "loss": 0.0526, "step": 9940 }, { "epoch": 0.1988329802965509, "grad_norm": 0.3458362817764282, "learning_rate": 9.99594381711321e-05, "loss": 0.0638, "step": 9950 }, { "epoch": 0.19903281243755244, "grad_norm": 0.17847840487957, "learning_rate": 9.995902092812358e-05, "loss": 0.0517, "step": 9960 }, { "epoch": 0.19923264457855402, "grad_norm": 0.22133462131023407, "learning_rate": 9.995860155095746e-05, "loss": 0.0558, "step": 9970 }, { "epoch": 0.19943247671955558, "grad_norm": 0.1805785447359085, "learning_rate": 9.995818003965168e-05, "loss": 0.0519, "step": 9980 }, { "epoch": 0.19963230886055713, "grad_norm": 0.23682667315006256, "learning_rate": 9.99577563942242e-05, "loss": 0.0567, "step": 9990 }, { "epoch": 0.1998321410015587, "grad_norm": 0.1721353679895401, "learning_rate": 9.995733061469315e-05, "loss": 0.0555, "step": 10000 }, { "epoch": 0.20003197314256024, "grad_norm": 0.1710163652896881, "learning_rate": 9.995690270107672e-05, "loss": 0.0497, "step": 10010 }, { "epoch": 0.2002318052835618, "grad_norm": 0.1524798721075058, "learning_rate": 9.995647265339318e-05, "loss": 0.06, "step": 10020 }, { "epoch": 0.20043163742456338, "grad_norm": 0.15767036378383636, "learning_rate": 9.99560404716609e-05, "loss": 0.0524, "step": 10030 }, { "epoch": 0.20063146956556493, "grad_norm": 0.18673346936702728, "learning_rate": 9.995560615589833e-05, "loss": 0.0564, "step": 10040 }, { "epoch": 0.2008313017065665, "grad_norm": 0.1259276419878006, "learning_rate": 9.995516970612405e-05, "loss": 0.0529, "step": 10050 }, { "epoch": 0.20103113384756804, "grad_norm": 0.15045855939388275, "learning_rate": 9.995473112235667e-05, "loss": 0.0535, "step": 10060 }, { "epoch": 0.2012309659885696, "grad_norm": 0.15967877209186554, "learning_rate": 9.995429040461497e-05, "loss": 0.0501, "step": 10070 }, { "epoch": 0.20143079812957115, "grad_norm": 0.20284846425056458, "learning_rate": 9.995384755291775e-05, "loss": 0.0543, "step": 10080 }, { "epoch": 0.2016306302705727, "grad_norm": 0.15686257183551788, "learning_rate": 9.995340256728392e-05, "loss": 0.0566, "step": 10090 }, { "epoch": 0.2018304624115743, "grad_norm": 0.1565326452255249, "learning_rate": 9.99529554477325e-05, "loss": 0.0527, "step": 10100 }, { "epoch": 0.20203029455257585, "grad_norm": 0.1345410794019699, "learning_rate": 9.99525061942826e-05, "loss": 0.0501, "step": 10110 }, { "epoch": 0.2022301266935774, "grad_norm": 0.1520226001739502, "learning_rate": 9.99520548069534e-05, "loss": 0.058, "step": 10120 }, { "epoch": 0.20242995883457895, "grad_norm": 0.1760067641735077, "learning_rate": 9.995160128576417e-05, "loss": 0.0468, "step": 10130 }, { "epoch": 0.2026297909755805, "grad_norm": 0.17050476372241974, "learning_rate": 9.995114563073431e-05, "loss": 0.0432, "step": 10140 }, { "epoch": 0.20282962311658206, "grad_norm": 0.15185725688934326, "learning_rate": 9.995068784188327e-05, "loss": 0.0538, "step": 10150 }, { "epoch": 0.20302945525758362, "grad_norm": 0.1205834448337555, "learning_rate": 9.99502279192306e-05, "loss": 0.053, "step": 10160 }, { "epoch": 0.2032292873985852, "grad_norm": 0.19883689284324646, "learning_rate": 9.994976586279597e-05, "loss": 0.0536, "step": 10170 }, { "epoch": 0.20342911953958676, "grad_norm": 0.19472934305667877, "learning_rate": 9.994930167259908e-05, "loss": 0.0574, "step": 10180 }, { "epoch": 0.2036289516805883, "grad_norm": 0.18957749009132385, "learning_rate": 9.994883534865979e-05, "loss": 0.0628, "step": 10190 }, { "epoch": 0.20382878382158987, "grad_norm": 0.13611629605293274, "learning_rate": 9.994836689099802e-05, "loss": 0.0502, "step": 10200 }, { "epoch": 0.20402861596259142, "grad_norm": 0.14873740077018738, "learning_rate": 9.994789629963378e-05, "loss": 0.0519, "step": 10210 }, { "epoch": 0.20422844810359297, "grad_norm": 0.13838165998458862, "learning_rate": 9.994742357458716e-05, "loss": 0.0507, "step": 10220 }, { "epoch": 0.20442828024459453, "grad_norm": 0.16728848218917847, "learning_rate": 9.994694871587837e-05, "loss": 0.067, "step": 10230 }, { "epoch": 0.2046281123855961, "grad_norm": 0.12920112907886505, "learning_rate": 9.994647172352767e-05, "loss": 0.0553, "step": 10240 }, { "epoch": 0.20482794452659767, "grad_norm": 0.18301625549793243, "learning_rate": 9.994599259755547e-05, "loss": 0.0452, "step": 10250 }, { "epoch": 0.20502777666759922, "grad_norm": 0.20114266872406006, "learning_rate": 9.99455113379822e-05, "loss": 0.0519, "step": 10260 }, { "epoch": 0.20522760880860078, "grad_norm": 0.18384051322937012, "learning_rate": 9.994502794482845e-05, "loss": 0.0477, "step": 10270 }, { "epoch": 0.20542744094960233, "grad_norm": 0.2155357152223587, "learning_rate": 9.994454241811486e-05, "loss": 0.0541, "step": 10280 }, { "epoch": 0.20562727309060388, "grad_norm": 0.18618622422218323, "learning_rate": 9.994405475786216e-05, "loss": 0.0458, "step": 10290 }, { "epoch": 0.20582710523160544, "grad_norm": 0.17016743123531342, "learning_rate": 9.994356496409119e-05, "loss": 0.0545, "step": 10300 }, { "epoch": 0.20602693737260702, "grad_norm": 0.19511985778808594, "learning_rate": 9.994307303682289e-05, "loss": 0.0552, "step": 10310 }, { "epoch": 0.20622676951360858, "grad_norm": 0.1623922735452652, "learning_rate": 9.994257897607825e-05, "loss": 0.0553, "step": 10320 }, { "epoch": 0.20642660165461013, "grad_norm": 0.17844389379024506, "learning_rate": 9.994208278187837e-05, "loss": 0.0426, "step": 10330 }, { "epoch": 0.20662643379561169, "grad_norm": 0.14061753451824188, "learning_rate": 9.994158445424448e-05, "loss": 0.0458, "step": 10340 }, { "epoch": 0.20682626593661324, "grad_norm": 0.1292824149131775, "learning_rate": 9.994108399319784e-05, "loss": 0.0509, "step": 10350 }, { "epoch": 0.2070260980776148, "grad_norm": 0.2176661193370819, "learning_rate": 9.994058139875983e-05, "loss": 0.0504, "step": 10360 }, { "epoch": 0.20722593021861635, "grad_norm": 0.20441779494285583, "learning_rate": 9.994007667095194e-05, "loss": 0.0481, "step": 10370 }, { "epoch": 0.20742576235961793, "grad_norm": 0.1715855747461319, "learning_rate": 9.993956980979572e-05, "loss": 0.062, "step": 10380 }, { "epoch": 0.2076255945006195, "grad_norm": 0.1884700208902359, "learning_rate": 9.993906081531281e-05, "loss": 0.0607, "step": 10390 }, { "epoch": 0.20782542664162104, "grad_norm": 0.16308657824993134, "learning_rate": 9.993854968752497e-05, "loss": 0.0545, "step": 10400 }, { "epoch": 0.2080252587826226, "grad_norm": 0.2434471845626831, "learning_rate": 9.993803642645401e-05, "loss": 0.053, "step": 10410 }, { "epoch": 0.20822509092362415, "grad_norm": 0.19018015265464783, "learning_rate": 9.993752103212188e-05, "loss": 0.0541, "step": 10420 }, { "epoch": 0.2084249230646257, "grad_norm": 0.20936784148216248, "learning_rate": 9.99370035045506e-05, "loss": 0.0571, "step": 10430 }, { "epoch": 0.20862475520562726, "grad_norm": 0.1633637547492981, "learning_rate": 9.993648384376227e-05, "loss": 0.0538, "step": 10440 }, { "epoch": 0.20882458734662884, "grad_norm": 0.16378900408744812, "learning_rate": 9.993596204977906e-05, "loss": 0.0561, "step": 10450 }, { "epoch": 0.2090244194876304, "grad_norm": 0.21289171278476715, "learning_rate": 9.993543812262332e-05, "loss": 0.0519, "step": 10460 }, { "epoch": 0.20922425162863195, "grad_norm": 0.13022556900978088, "learning_rate": 9.993491206231738e-05, "loss": 0.0462, "step": 10470 }, { "epoch": 0.2094240837696335, "grad_norm": 0.18391945958137512, "learning_rate": 9.993438386888372e-05, "loss": 0.0517, "step": 10480 }, { "epoch": 0.20962391591063506, "grad_norm": 0.17387184500694275, "learning_rate": 9.993385354234492e-05, "loss": 0.0461, "step": 10490 }, { "epoch": 0.20982374805163662, "grad_norm": 0.1628747284412384, "learning_rate": 9.993332108272363e-05, "loss": 0.053, "step": 10500 }, { "epoch": 0.21002358019263817, "grad_norm": 0.15129905939102173, "learning_rate": 9.993278649004261e-05, "loss": 0.0512, "step": 10510 }, { "epoch": 0.21022341233363975, "grad_norm": 0.1762315034866333, "learning_rate": 9.993224976432466e-05, "loss": 0.0573, "step": 10520 }, { "epoch": 0.2104232444746413, "grad_norm": 0.17189140617847443, "learning_rate": 9.993171090559275e-05, "loss": 0.0603, "step": 10530 }, { "epoch": 0.21062307661564286, "grad_norm": 0.1605214625597, "learning_rate": 9.993116991386986e-05, "loss": 0.047, "step": 10540 }, { "epoch": 0.21082290875664442, "grad_norm": 0.13825437426567078, "learning_rate": 9.993062678917913e-05, "loss": 0.0637, "step": 10550 }, { "epoch": 0.21102274089764597, "grad_norm": 0.16141779720783234, "learning_rate": 9.993008153154372e-05, "loss": 0.0488, "step": 10560 }, { "epoch": 0.21122257303864753, "grad_norm": 0.15583735704421997, "learning_rate": 9.992953414098699e-05, "loss": 0.0567, "step": 10570 }, { "epoch": 0.21142240517964908, "grad_norm": 0.1514456570148468, "learning_rate": 9.992898461753226e-05, "loss": 0.0516, "step": 10580 }, { "epoch": 0.21162223732065066, "grad_norm": 0.1567309945821762, "learning_rate": 9.992843296120305e-05, "loss": 0.0497, "step": 10590 }, { "epoch": 0.21182206946165222, "grad_norm": 0.11032943427562714, "learning_rate": 9.992787917202288e-05, "loss": 0.0521, "step": 10600 }, { "epoch": 0.21202190160265377, "grad_norm": 0.15139853954315186, "learning_rate": 9.992732325001545e-05, "loss": 0.0536, "step": 10610 }, { "epoch": 0.21222173374365533, "grad_norm": 0.1425962895154953, "learning_rate": 9.992676519520447e-05, "loss": 0.0509, "step": 10620 }, { "epoch": 0.21242156588465688, "grad_norm": 0.18740779161453247, "learning_rate": 9.992620500761382e-05, "loss": 0.0543, "step": 10630 }, { "epoch": 0.21262139802565844, "grad_norm": 0.13504067063331604, "learning_rate": 9.992564268726741e-05, "loss": 0.052, "step": 10640 }, { "epoch": 0.21282123016666002, "grad_norm": 0.17989970743656158, "learning_rate": 9.992507823418924e-05, "loss": 0.0507, "step": 10650 }, { "epoch": 0.21302106230766157, "grad_norm": 0.1268872767686844, "learning_rate": 9.992451164840347e-05, "loss": 0.0487, "step": 10660 }, { "epoch": 0.21322089444866313, "grad_norm": 0.1595105081796646, "learning_rate": 9.992394292993426e-05, "loss": 0.0568, "step": 10670 }, { "epoch": 0.21342072658966468, "grad_norm": 0.1454349309206009, "learning_rate": 9.992337207880593e-05, "loss": 0.0549, "step": 10680 }, { "epoch": 0.21362055873066624, "grad_norm": 0.1306333988904953, "learning_rate": 9.992279909504285e-05, "loss": 0.0465, "step": 10690 }, { "epoch": 0.2138203908716678, "grad_norm": 0.1280200332403183, "learning_rate": 9.992222397866951e-05, "loss": 0.0526, "step": 10700 }, { "epoch": 0.21402022301266935, "grad_norm": 0.163410484790802, "learning_rate": 9.992164672971046e-05, "loss": 0.0586, "step": 10710 }, { "epoch": 0.21422005515367093, "grad_norm": 0.122348852455616, "learning_rate": 9.992106734819037e-05, "loss": 0.049, "step": 10720 }, { "epoch": 0.21441988729467248, "grad_norm": 0.15395112335681915, "learning_rate": 9.992048583413401e-05, "loss": 0.0544, "step": 10730 }, { "epoch": 0.21461971943567404, "grad_norm": 0.1181684285402298, "learning_rate": 9.99199021875662e-05, "loss": 0.0575, "step": 10740 }, { "epoch": 0.2148195515766756, "grad_norm": 0.19219717383384705, "learning_rate": 9.991931640851185e-05, "loss": 0.053, "step": 10750 }, { "epoch": 0.21501938371767715, "grad_norm": 0.17827177047729492, "learning_rate": 9.991872849699603e-05, "loss": 0.0498, "step": 10760 }, { "epoch": 0.2152192158586787, "grad_norm": 0.16559326648712158, "learning_rate": 9.991813845304383e-05, "loss": 0.0497, "step": 10770 }, { "epoch": 0.21541904799968026, "grad_norm": 0.1743195652961731, "learning_rate": 9.991754627668044e-05, "loss": 0.0528, "step": 10780 }, { "epoch": 0.21561888014068184, "grad_norm": 0.18648523092269897, "learning_rate": 9.99169519679312e-05, "loss": 0.0485, "step": 10790 }, { "epoch": 0.2158187122816834, "grad_norm": 0.14329758286476135, "learning_rate": 9.991635552682145e-05, "loss": 0.0572, "step": 10800 }, { "epoch": 0.21601854442268495, "grad_norm": 0.14493106305599213, "learning_rate": 9.991575695337671e-05, "loss": 0.052, "step": 10810 }, { "epoch": 0.2162183765636865, "grad_norm": 0.18460561335086823, "learning_rate": 9.991515624762251e-05, "loss": 0.0527, "step": 10820 }, { "epoch": 0.21641820870468806, "grad_norm": 0.20318512618541718, "learning_rate": 9.991455340958455e-05, "loss": 0.0519, "step": 10830 }, { "epoch": 0.2166180408456896, "grad_norm": 0.19959454238414764, "learning_rate": 9.991394843928856e-05, "loss": 0.049, "step": 10840 }, { "epoch": 0.21681787298669117, "grad_norm": 0.16077440977096558, "learning_rate": 9.991334133676039e-05, "loss": 0.051, "step": 10850 }, { "epoch": 0.21701770512769275, "grad_norm": 0.18748839199543, "learning_rate": 9.991273210202596e-05, "loss": 0.0562, "step": 10860 }, { "epoch": 0.2172175372686943, "grad_norm": 0.1511429250240326, "learning_rate": 9.991212073511132e-05, "loss": 0.0472, "step": 10870 }, { "epoch": 0.21741736940969586, "grad_norm": 0.16991844773292542, "learning_rate": 9.991150723604257e-05, "loss": 0.0522, "step": 10880 }, { "epoch": 0.21761720155069741, "grad_norm": 0.15815606713294983, "learning_rate": 9.991089160484593e-05, "loss": 0.0614, "step": 10890 }, { "epoch": 0.21781703369169897, "grad_norm": 0.2248021364212036, "learning_rate": 9.991027384154767e-05, "loss": 0.054, "step": 10900 }, { "epoch": 0.21801686583270052, "grad_norm": 0.1782161146402359, "learning_rate": 9.990965394617422e-05, "loss": 0.0405, "step": 10910 }, { "epoch": 0.21821669797370208, "grad_norm": 0.13456417620182037, "learning_rate": 9.990903191875203e-05, "loss": 0.0564, "step": 10920 }, { "epoch": 0.21841653011470366, "grad_norm": 0.21069969236850739, "learning_rate": 9.990840775930768e-05, "loss": 0.0508, "step": 10930 }, { "epoch": 0.21861636225570522, "grad_norm": 0.12548571825027466, "learning_rate": 9.990778146786784e-05, "loss": 0.0476, "step": 10940 }, { "epoch": 0.21881619439670677, "grad_norm": 0.16942986845970154, "learning_rate": 9.990715304445926e-05, "loss": 0.0464, "step": 10950 }, { "epoch": 0.21901602653770832, "grad_norm": 0.15509551763534546, "learning_rate": 9.990652248910877e-05, "loss": 0.054, "step": 10960 }, { "epoch": 0.21921585867870988, "grad_norm": 0.16650588810443878, "learning_rate": 9.990588980184333e-05, "loss": 0.0487, "step": 10970 }, { "epoch": 0.21941569081971143, "grad_norm": 0.14600111544132233, "learning_rate": 9.990525498268996e-05, "loss": 0.0438, "step": 10980 }, { "epoch": 0.219615522960713, "grad_norm": 0.12299957126379013, "learning_rate": 9.990461803167577e-05, "loss": 0.049, "step": 10990 }, { "epoch": 0.21981535510171457, "grad_norm": 0.14967553317546844, "learning_rate": 9.990397894882799e-05, "loss": 0.0461, "step": 11000 }, { "epoch": 0.22001518724271613, "grad_norm": 0.13387364149093628, "learning_rate": 9.99033377341739e-05, "loss": 0.049, "step": 11010 }, { "epoch": 0.22021501938371768, "grad_norm": 0.1689067780971527, "learning_rate": 9.990269438774089e-05, "loss": 0.0529, "step": 11020 }, { "epoch": 0.22041485152471924, "grad_norm": 0.18959443271160126, "learning_rate": 9.990204890955645e-05, "loss": 0.0447, "step": 11030 }, { "epoch": 0.2206146836657208, "grad_norm": 0.15619084239006042, "learning_rate": 9.990140129964815e-05, "loss": 0.0447, "step": 11040 }, { "epoch": 0.22081451580672234, "grad_norm": 0.18354254961013794, "learning_rate": 9.990075155804368e-05, "loss": 0.0567, "step": 11050 }, { "epoch": 0.2210143479477239, "grad_norm": 0.17371031641960144, "learning_rate": 9.990009968477075e-05, "loss": 0.0574, "step": 11060 }, { "epoch": 0.22121418008872548, "grad_norm": 0.16731469333171844, "learning_rate": 9.989944567985725e-05, "loss": 0.0519, "step": 11070 }, { "epoch": 0.22141401222972704, "grad_norm": 0.1481328308582306, "learning_rate": 9.989878954333109e-05, "loss": 0.0477, "step": 11080 }, { "epoch": 0.2216138443707286, "grad_norm": 0.14579632878303528, "learning_rate": 9.989813127522029e-05, "loss": 0.0452, "step": 11090 }, { "epoch": 0.22181367651173015, "grad_norm": 0.18501797318458557, "learning_rate": 9.989747087555301e-05, "loss": 0.0484, "step": 11100 }, { "epoch": 0.2220135086527317, "grad_norm": 0.12208293378353119, "learning_rate": 9.989680834435744e-05, "loss": 0.0522, "step": 11110 }, { "epoch": 0.22221334079373325, "grad_norm": 0.17203949391841888, "learning_rate": 9.989614368166187e-05, "loss": 0.0507, "step": 11120 }, { "epoch": 0.2224131729347348, "grad_norm": 0.2157391458749771, "learning_rate": 9.989547688749472e-05, "loss": 0.0485, "step": 11130 }, { "epoch": 0.2226130050757364, "grad_norm": 0.15918953716754913, "learning_rate": 9.989480796188444e-05, "loss": 0.0471, "step": 11140 }, { "epoch": 0.22281283721673795, "grad_norm": 0.13839900493621826, "learning_rate": 9.989413690485963e-05, "loss": 0.05, "step": 11150 }, { "epoch": 0.2230126693577395, "grad_norm": 0.19680099189281464, "learning_rate": 9.989346371644895e-05, "loss": 0.0573, "step": 11160 }, { "epoch": 0.22321250149874106, "grad_norm": 0.12995538115501404, "learning_rate": 9.989278839668116e-05, "loss": 0.0512, "step": 11170 }, { "epoch": 0.2234123336397426, "grad_norm": 0.16585160791873932, "learning_rate": 9.98921109455851e-05, "loss": 0.05, "step": 11180 }, { "epoch": 0.22361216578074417, "grad_norm": 0.17990340292453766, "learning_rate": 9.989143136318973e-05, "loss": 0.0576, "step": 11190 }, { "epoch": 0.22381199792174572, "grad_norm": 0.17364893853664398, "learning_rate": 9.989074964952405e-05, "loss": 0.0498, "step": 11200 }, { "epoch": 0.2240118300627473, "grad_norm": 0.20412075519561768, "learning_rate": 9.989006580461719e-05, "loss": 0.0658, "step": 11210 }, { "epoch": 0.22421166220374886, "grad_norm": 0.15038953721523285, "learning_rate": 9.988937982849838e-05, "loss": 0.0457, "step": 11220 }, { "epoch": 0.2244114943447504, "grad_norm": 0.18005400896072388, "learning_rate": 9.988869172119693e-05, "loss": 0.0585, "step": 11230 }, { "epoch": 0.22461132648575197, "grad_norm": 0.2202436625957489, "learning_rate": 9.98880014827422e-05, "loss": 0.0515, "step": 11240 }, { "epoch": 0.22481115862675352, "grad_norm": 0.17608432471752167, "learning_rate": 9.988730911316368e-05, "loss": 0.05, "step": 11250 }, { "epoch": 0.22501099076775508, "grad_norm": 0.2403896152973175, "learning_rate": 9.988661461249099e-05, "loss": 0.0447, "step": 11260 }, { "epoch": 0.22521082290875666, "grad_norm": 0.17000557482242584, "learning_rate": 9.988591798075375e-05, "loss": 0.0466, "step": 11270 }, { "epoch": 0.2254106550497582, "grad_norm": 0.21063941717147827, "learning_rate": 9.988521921798175e-05, "loss": 0.0589, "step": 11280 }, { "epoch": 0.22561048719075977, "grad_norm": 0.14422529935836792, "learning_rate": 9.988451832420483e-05, "loss": 0.0437, "step": 11290 }, { "epoch": 0.22581031933176132, "grad_norm": 0.19333526492118835, "learning_rate": 9.98838152994529e-05, "loss": 0.0436, "step": 11300 }, { "epoch": 0.22601015147276288, "grad_norm": 0.14590537548065186, "learning_rate": 9.988311014375605e-05, "loss": 0.0483, "step": 11310 }, { "epoch": 0.22620998361376443, "grad_norm": 0.20001138746738434, "learning_rate": 9.988240285714435e-05, "loss": 0.0569, "step": 11320 }, { "epoch": 0.226409815754766, "grad_norm": 0.13148988783359528, "learning_rate": 9.988169343964804e-05, "loss": 0.0546, "step": 11330 }, { "epoch": 0.22660964789576757, "grad_norm": 0.17364567518234253, "learning_rate": 9.988098189129745e-05, "loss": 0.0491, "step": 11340 }, { "epoch": 0.22680948003676912, "grad_norm": 0.1497020274400711, "learning_rate": 9.988026821212292e-05, "loss": 0.0471, "step": 11350 }, { "epoch": 0.22700931217777068, "grad_norm": 0.16975797712802887, "learning_rate": 9.987955240215497e-05, "loss": 0.0525, "step": 11360 }, { "epoch": 0.22720914431877223, "grad_norm": 0.14490382373332977, "learning_rate": 9.987883446142418e-05, "loss": 0.0509, "step": 11370 }, { "epoch": 0.2274089764597738, "grad_norm": 0.16125303506851196, "learning_rate": 9.98781143899612e-05, "loss": 0.0659, "step": 11380 }, { "epoch": 0.22760880860077534, "grad_norm": 0.15745235979557037, "learning_rate": 9.987739218779681e-05, "loss": 0.0477, "step": 11390 }, { "epoch": 0.2278086407417769, "grad_norm": 0.18695437908172607, "learning_rate": 9.987666785496186e-05, "loss": 0.0414, "step": 11400 }, { "epoch": 0.22800847288277848, "grad_norm": 0.21234610676765442, "learning_rate": 9.987594139148728e-05, "loss": 0.0596, "step": 11410 }, { "epoch": 0.22820830502378003, "grad_norm": 0.15894265472888947, "learning_rate": 9.98752127974041e-05, "loss": 0.0522, "step": 11420 }, { "epoch": 0.2284081371647816, "grad_norm": 0.227457657456398, "learning_rate": 9.987448207274346e-05, "loss": 0.0562, "step": 11430 }, { "epoch": 0.22860796930578314, "grad_norm": 0.19073955714702606, "learning_rate": 9.987374921753657e-05, "loss": 0.0571, "step": 11440 }, { "epoch": 0.2288078014467847, "grad_norm": 0.1858953982591629, "learning_rate": 9.987301423181473e-05, "loss": 0.0618, "step": 11450 }, { "epoch": 0.22900763358778625, "grad_norm": 0.16771988570690155, "learning_rate": 9.987227711560934e-05, "loss": 0.0579, "step": 11460 }, { "epoch": 0.2292074657287878, "grad_norm": 0.14603176712989807, "learning_rate": 9.98715378689519e-05, "loss": 0.045, "step": 11470 }, { "epoch": 0.2294072978697894, "grad_norm": 0.1701815277338028, "learning_rate": 9.987079649187398e-05, "loss": 0.0501, "step": 11480 }, { "epoch": 0.22960713001079094, "grad_norm": 0.20592747628688812, "learning_rate": 9.987005298440723e-05, "loss": 0.0485, "step": 11490 }, { "epoch": 0.2298069621517925, "grad_norm": 0.19840696454048157, "learning_rate": 9.986930734658346e-05, "loss": 0.0524, "step": 11500 }, { "epoch": 0.23000679429279405, "grad_norm": 0.11839913576841354, "learning_rate": 9.986855957843448e-05, "loss": 0.0471, "step": 11510 }, { "epoch": 0.2302066264337956, "grad_norm": 0.16156554222106934, "learning_rate": 9.986780967999223e-05, "loss": 0.0548, "step": 11520 }, { "epoch": 0.23040645857479716, "grad_norm": 0.1845076084136963, "learning_rate": 9.986705765128878e-05, "loss": 0.0545, "step": 11530 }, { "epoch": 0.23060629071579872, "grad_norm": 0.15438076853752136, "learning_rate": 9.986630349235623e-05, "loss": 0.0484, "step": 11540 }, { "epoch": 0.2308061228568003, "grad_norm": 0.1704416424036026, "learning_rate": 9.98655472032268e-05, "loss": 0.0505, "step": 11550 }, { "epoch": 0.23100595499780185, "grad_norm": 0.1612798422574997, "learning_rate": 9.98647887839328e-05, "loss": 0.058, "step": 11560 }, { "epoch": 0.2312057871388034, "grad_norm": 0.14581964910030365, "learning_rate": 9.986402823450663e-05, "loss": 0.0443, "step": 11570 }, { "epoch": 0.23140561927980496, "grad_norm": 0.1491188108921051, "learning_rate": 9.986326555498077e-05, "loss": 0.0552, "step": 11580 }, { "epoch": 0.23160545142080652, "grad_norm": 0.13303197920322418, "learning_rate": 9.986250074538782e-05, "loss": 0.049, "step": 11590 }, { "epoch": 0.23180528356180807, "grad_norm": 0.21345168352127075, "learning_rate": 9.986173380576043e-05, "loss": 0.0513, "step": 11600 }, { "epoch": 0.23200511570280963, "grad_norm": 0.15327788889408112, "learning_rate": 9.986096473613138e-05, "loss": 0.0468, "step": 11610 }, { "epoch": 0.2322049478438112, "grad_norm": 0.24058538675308228, "learning_rate": 9.98601935365335e-05, "loss": 0.051, "step": 11620 }, { "epoch": 0.23240477998481276, "grad_norm": 0.13964585959911346, "learning_rate": 9.985942020699975e-05, "loss": 0.0444, "step": 11630 }, { "epoch": 0.23260461212581432, "grad_norm": 0.1836652308702469, "learning_rate": 9.985864474756316e-05, "loss": 0.0535, "step": 11640 }, { "epoch": 0.23280444426681587, "grad_norm": 0.17092560231685638, "learning_rate": 9.985786715825688e-05, "loss": 0.0561, "step": 11650 }, { "epoch": 0.23300427640781743, "grad_norm": 0.16557812690734863, "learning_rate": 9.98570874391141e-05, "loss": 0.0567, "step": 11660 }, { "epoch": 0.23320410854881898, "grad_norm": 0.17623217403888702, "learning_rate": 9.985630559016812e-05, "loss": 0.0543, "step": 11670 }, { "epoch": 0.23340394068982054, "grad_norm": 0.1533610075712204, "learning_rate": 9.985552161145238e-05, "loss": 0.0524, "step": 11680 }, { "epoch": 0.23360377283082212, "grad_norm": 0.1698247343301773, "learning_rate": 9.985473550300032e-05, "loss": 0.059, "step": 11690 }, { "epoch": 0.23380360497182368, "grad_norm": 0.13597367703914642, "learning_rate": 9.985394726484555e-05, "loss": 0.0553, "step": 11700 }, { "epoch": 0.23400343711282523, "grad_norm": 0.392443984746933, "learning_rate": 9.985315689702175e-05, "loss": 0.0504, "step": 11710 }, { "epoch": 0.23420326925382678, "grad_norm": 0.1607077270746231, "learning_rate": 9.985236439956265e-05, "loss": 0.053, "step": 11720 }, { "epoch": 0.23440310139482834, "grad_norm": 0.13663797080516815, "learning_rate": 9.985156977250215e-05, "loss": 0.0521, "step": 11730 }, { "epoch": 0.2346029335358299, "grad_norm": 0.14092139899730682, "learning_rate": 9.985077301587415e-05, "loss": 0.0496, "step": 11740 }, { "epoch": 0.23480276567683145, "grad_norm": 0.12238074839115143, "learning_rate": 9.984997412971271e-05, "loss": 0.0563, "step": 11750 }, { "epoch": 0.23500259781783303, "grad_norm": 0.16750532388687134, "learning_rate": 9.984917311405196e-05, "loss": 0.0675, "step": 11760 }, { "epoch": 0.23520242995883459, "grad_norm": 0.20631128549575806, "learning_rate": 9.98483699689261e-05, "loss": 0.0514, "step": 11770 }, { "epoch": 0.23540226209983614, "grad_norm": 0.20816092193126678, "learning_rate": 9.984756469436944e-05, "loss": 0.0489, "step": 11780 }, { "epoch": 0.2356020942408377, "grad_norm": 0.16802768409252167, "learning_rate": 9.984675729041642e-05, "loss": 0.0485, "step": 11790 }, { "epoch": 0.23580192638183925, "grad_norm": 0.1842728555202484, "learning_rate": 9.984594775710148e-05, "loss": 0.0533, "step": 11800 }, { "epoch": 0.2360017585228408, "grad_norm": 0.19958104193210602, "learning_rate": 9.984513609445922e-05, "loss": 0.053, "step": 11810 }, { "epoch": 0.23620159066384236, "grad_norm": 0.13094079494476318, "learning_rate": 9.984432230252432e-05, "loss": 0.0509, "step": 11820 }, { "epoch": 0.23640142280484394, "grad_norm": 0.17528218030929565, "learning_rate": 9.984350638133153e-05, "loss": 0.0531, "step": 11830 }, { "epoch": 0.2366012549458455, "grad_norm": 0.2073843777179718, "learning_rate": 9.984268833091573e-05, "loss": 0.0532, "step": 11840 }, { "epoch": 0.23680108708684705, "grad_norm": 0.18830768764019012, "learning_rate": 9.984186815131182e-05, "loss": 0.0499, "step": 11850 }, { "epoch": 0.2370009192278486, "grad_norm": 0.16763773560523987, "learning_rate": 9.98410458425549e-05, "loss": 0.0482, "step": 11860 }, { "epoch": 0.23720075136885016, "grad_norm": 0.23877620697021484, "learning_rate": 9.984022140468004e-05, "loss": 0.0505, "step": 11870 }, { "epoch": 0.23740058350985171, "grad_norm": 0.1566130518913269, "learning_rate": 9.983939483772248e-05, "loss": 0.0491, "step": 11880 }, { "epoch": 0.23760041565085327, "grad_norm": 0.15757641196250916, "learning_rate": 9.983856614171753e-05, "loss": 0.0491, "step": 11890 }, { "epoch": 0.23780024779185485, "grad_norm": 0.16664496064186096, "learning_rate": 9.983773531670061e-05, "loss": 0.0489, "step": 11900 }, { "epoch": 0.2380000799328564, "grad_norm": 0.12951576709747314, "learning_rate": 9.983690236270718e-05, "loss": 0.0474, "step": 11910 }, { "epoch": 0.23819991207385796, "grad_norm": 0.1705942004919052, "learning_rate": 9.983606727977284e-05, "loss": 0.0499, "step": 11920 }, { "epoch": 0.23839974421485952, "grad_norm": 0.13685102760791779, "learning_rate": 9.983523006793325e-05, "loss": 0.0451, "step": 11930 }, { "epoch": 0.23859957635586107, "grad_norm": 0.2021612673997879, "learning_rate": 9.983439072722418e-05, "loss": 0.0457, "step": 11940 }, { "epoch": 0.23879940849686263, "grad_norm": 0.17946173250675201, "learning_rate": 9.98335492576815e-05, "loss": 0.0481, "step": 11950 }, { "epoch": 0.2389992406378642, "grad_norm": 0.1561555713415146, "learning_rate": 9.983270565934113e-05, "loss": 0.0568, "step": 11960 }, { "epoch": 0.23919907277886576, "grad_norm": 0.15384316444396973, "learning_rate": 9.983185993223912e-05, "loss": 0.0431, "step": 11970 }, { "epoch": 0.23939890491986732, "grad_norm": 0.17759539186954498, "learning_rate": 9.983101207641161e-05, "loss": 0.0516, "step": 11980 }, { "epoch": 0.23959873706086887, "grad_norm": 0.1817239373922348, "learning_rate": 9.983016209189478e-05, "loss": 0.0522, "step": 11990 }, { "epoch": 0.23979856920187043, "grad_norm": 0.44483181834220886, "learning_rate": 9.982930997872499e-05, "loss": 0.0552, "step": 12000 }, { "epoch": 0.23999840134287198, "grad_norm": 0.23219752311706543, "learning_rate": 9.982845573693862e-05, "loss": 0.0427, "step": 12010 }, { "epoch": 0.24019823348387354, "grad_norm": 0.15880756080150604, "learning_rate": 9.982759936657214e-05, "loss": 0.0508, "step": 12020 }, { "epoch": 0.24039806562487512, "grad_norm": 0.18755903840065002, "learning_rate": 9.982674086766216e-05, "loss": 0.0509, "step": 12030 }, { "epoch": 0.24059789776587667, "grad_norm": 0.2154257893562317, "learning_rate": 9.982588024024535e-05, "loss": 0.0491, "step": 12040 }, { "epoch": 0.24079772990687823, "grad_norm": 0.17034173011779785, "learning_rate": 9.982501748435846e-05, "loss": 0.0495, "step": 12050 }, { "epoch": 0.24099756204787978, "grad_norm": 0.13413815200328827, "learning_rate": 9.982415260003837e-05, "loss": 0.0465, "step": 12060 }, { "epoch": 0.24119739418888134, "grad_norm": 0.2093474417924881, "learning_rate": 9.982328558732199e-05, "loss": 0.058, "step": 12070 }, { "epoch": 0.2413972263298829, "grad_norm": 0.1368514895439148, "learning_rate": 9.982241644624639e-05, "loss": 0.0463, "step": 12080 }, { "epoch": 0.24159705847088445, "grad_norm": 0.16664262115955353, "learning_rate": 9.982154517684868e-05, "loss": 0.0468, "step": 12090 }, { "epoch": 0.24179689061188603, "grad_norm": 0.17312853038311005, "learning_rate": 9.98206717791661e-05, "loss": 0.0432, "step": 12100 }, { "epoch": 0.24199672275288758, "grad_norm": 0.20208974182605743, "learning_rate": 9.981979625323593e-05, "loss": 0.0466, "step": 12110 }, { "epoch": 0.24219655489388914, "grad_norm": 0.12522558867931366, "learning_rate": 9.981891859909559e-05, "loss": 0.052, "step": 12120 }, { "epoch": 0.2423963870348907, "grad_norm": 0.14005134999752045, "learning_rate": 9.981803881678258e-05, "loss": 0.0485, "step": 12130 }, { "epoch": 0.24259621917589225, "grad_norm": 0.17289189994335175, "learning_rate": 9.981715690633444e-05, "loss": 0.0525, "step": 12140 }, { "epoch": 0.2427960513168938, "grad_norm": 0.15438681840896606, "learning_rate": 9.981627286778889e-05, "loss": 0.0542, "step": 12150 }, { "epoch": 0.24299588345789536, "grad_norm": 0.14909714460372925, "learning_rate": 9.98153867011837e-05, "loss": 0.0523, "step": 12160 }, { "epoch": 0.24319571559889694, "grad_norm": 0.2687230706214905, "learning_rate": 9.981449840655668e-05, "loss": 0.0443, "step": 12170 }, { "epoch": 0.2433955477398985, "grad_norm": 0.191701278090477, "learning_rate": 9.981360798394581e-05, "loss": 0.0552, "step": 12180 }, { "epoch": 0.24359537988090005, "grad_norm": 0.18313346803188324, "learning_rate": 9.98127154333891e-05, "loss": 0.0456, "step": 12190 }, { "epoch": 0.2437952120219016, "grad_norm": 0.14058764278888702, "learning_rate": 9.981182075492473e-05, "loss": 0.0472, "step": 12200 }, { "epoch": 0.24399504416290316, "grad_norm": 0.21919754147529602, "learning_rate": 9.981092394859085e-05, "loss": 0.0493, "step": 12210 }, { "epoch": 0.2441948763039047, "grad_norm": 0.13449420034885406, "learning_rate": 9.981002501442583e-05, "loss": 0.0463, "step": 12220 }, { "epoch": 0.24439470844490627, "grad_norm": 0.19462211430072784, "learning_rate": 9.980912395246802e-05, "loss": 0.0555, "step": 12230 }, { "epoch": 0.24459454058590785, "grad_norm": 0.20526784658432007, "learning_rate": 9.980822076275595e-05, "loss": 0.0528, "step": 12240 }, { "epoch": 0.2447943727269094, "grad_norm": 0.14470280706882477, "learning_rate": 9.98073154453282e-05, "loss": 0.0552, "step": 12250 }, { "epoch": 0.24499420486791096, "grad_norm": 0.17561262845993042, "learning_rate": 9.980640800022341e-05, "loss": 0.0511, "step": 12260 }, { "epoch": 0.2451940370089125, "grad_norm": 0.16772854328155518, "learning_rate": 9.980549842748038e-05, "loss": 0.0458, "step": 12270 }, { "epoch": 0.24539386914991407, "grad_norm": 0.15269288420677185, "learning_rate": 9.980458672713795e-05, "loss": 0.0454, "step": 12280 }, { "epoch": 0.24559370129091562, "grad_norm": 0.16385403275489807, "learning_rate": 9.980367289923508e-05, "loss": 0.0465, "step": 12290 }, { "epoch": 0.24579353343191718, "grad_norm": 0.138178288936615, "learning_rate": 9.980275694381078e-05, "loss": 0.0438, "step": 12300 }, { "epoch": 0.24599336557291876, "grad_norm": 0.16072142124176025, "learning_rate": 9.98018388609042e-05, "loss": 0.0471, "step": 12310 }, { "epoch": 0.24619319771392031, "grad_norm": 0.1598787158727646, "learning_rate": 9.980091865055457e-05, "loss": 0.0581, "step": 12320 }, { "epoch": 0.24639302985492187, "grad_norm": 0.11732111871242523, "learning_rate": 9.979999631280115e-05, "loss": 0.0489, "step": 12330 }, { "epoch": 0.24659286199592342, "grad_norm": 0.17563173174858093, "learning_rate": 9.97990718476834e-05, "loss": 0.0541, "step": 12340 }, { "epoch": 0.24679269413692498, "grad_norm": 0.14813585579395294, "learning_rate": 9.979814525524079e-05, "loss": 0.0553, "step": 12350 }, { "epoch": 0.24699252627792653, "grad_norm": 0.1759297251701355, "learning_rate": 9.979721653551286e-05, "loss": 0.0518, "step": 12360 }, { "epoch": 0.2471923584189281, "grad_norm": 0.15239134430885315, "learning_rate": 9.979628568853934e-05, "loss": 0.0539, "step": 12370 }, { "epoch": 0.24739219055992967, "grad_norm": 0.1721995621919632, "learning_rate": 9.979535271435997e-05, "loss": 0.0577, "step": 12380 }, { "epoch": 0.24759202270093122, "grad_norm": 0.14855334162712097, "learning_rate": 9.979441761301462e-05, "loss": 0.0555, "step": 12390 }, { "epoch": 0.24779185484193278, "grad_norm": 0.16792750358581543, "learning_rate": 9.979348038454323e-05, "loss": 0.0469, "step": 12400 }, { "epoch": 0.24799168698293433, "grad_norm": 0.15621371567249298, "learning_rate": 9.979254102898582e-05, "loss": 0.0485, "step": 12410 }, { "epoch": 0.2481915191239359, "grad_norm": 0.15978743135929108, "learning_rate": 9.979159954638253e-05, "loss": 0.0516, "step": 12420 }, { "epoch": 0.24839135126493744, "grad_norm": 0.18345585465431213, "learning_rate": 9.979065593677358e-05, "loss": 0.0455, "step": 12430 }, { "epoch": 0.248591183405939, "grad_norm": 0.16295945644378662, "learning_rate": 9.978971020019927e-05, "loss": 0.0519, "step": 12440 }, { "epoch": 0.24879101554694058, "grad_norm": 0.13439030945301056, "learning_rate": 9.978876233670002e-05, "loss": 0.0559, "step": 12450 }, { "epoch": 0.24899084768794214, "grad_norm": 0.14263786375522614, "learning_rate": 9.978781234631628e-05, "loss": 0.046, "step": 12460 }, { "epoch": 0.2491906798289437, "grad_norm": 0.23314720392227173, "learning_rate": 9.97868602290887e-05, "loss": 0.0488, "step": 12470 }, { "epoch": 0.24939051196994524, "grad_norm": 0.1379610151052475, "learning_rate": 9.978590598505788e-05, "loss": 0.0523, "step": 12480 }, { "epoch": 0.2495903441109468, "grad_norm": 0.1507132202386856, "learning_rate": 9.978494961426463e-05, "loss": 0.0483, "step": 12490 }, { "epoch": 0.24979017625194835, "grad_norm": 0.1542244702577591, "learning_rate": 9.978399111674979e-05, "loss": 0.0535, "step": 12500 }, { "epoch": 0.2499900083929499, "grad_norm": 0.1801542341709137, "learning_rate": 9.978303049255431e-05, "loss": 0.0538, "step": 12510 }, { "epoch": 0.2501898405339515, "grad_norm": 0.15892423689365387, "learning_rate": 9.97820677417192e-05, "loss": 0.0518, "step": 12520 }, { "epoch": 0.25038967267495305, "grad_norm": 0.15451423823833466, "learning_rate": 9.978110286428564e-05, "loss": 0.054, "step": 12530 }, { "epoch": 0.2505895048159546, "grad_norm": 0.1611742228269577, "learning_rate": 9.978013586029479e-05, "loss": 0.0484, "step": 12540 }, { "epoch": 0.25078933695695615, "grad_norm": 0.17799244821071625, "learning_rate": 9.9779166729788e-05, "loss": 0.0562, "step": 12550 }, { "epoch": 0.2509891690979577, "grad_norm": 0.13966134190559387, "learning_rate": 9.977819547280665e-05, "loss": 0.052, "step": 12560 }, { "epoch": 0.25118900123895926, "grad_norm": 0.16549299657344818, "learning_rate": 9.977722208939224e-05, "loss": 0.0583, "step": 12570 }, { "epoch": 0.2513888333799608, "grad_norm": 0.12922939658164978, "learning_rate": 9.977624657958633e-05, "loss": 0.0482, "step": 12580 }, { "epoch": 0.2515886655209624, "grad_norm": 0.22786763310432434, "learning_rate": 9.977526894343061e-05, "loss": 0.045, "step": 12590 }, { "epoch": 0.25178849766196393, "grad_norm": 0.13180741667747498, "learning_rate": 9.977428918096685e-05, "loss": 0.0479, "step": 12600 }, { "epoch": 0.2519883298029655, "grad_norm": 0.20020592212677002, "learning_rate": 9.977330729223691e-05, "loss": 0.0493, "step": 12610 }, { "epoch": 0.2521881619439671, "grad_norm": 0.15291725099086761, "learning_rate": 9.97723232772827e-05, "loss": 0.0521, "step": 12620 }, { "epoch": 0.25238799408496865, "grad_norm": 0.17392800748348236, "learning_rate": 9.977133713614627e-05, "loss": 0.0394, "step": 12630 }, { "epoch": 0.2525878262259702, "grad_norm": 0.12706726789474487, "learning_rate": 9.977034886886976e-05, "loss": 0.0507, "step": 12640 }, { "epoch": 0.25278765836697176, "grad_norm": 0.16102571785449982, "learning_rate": 9.976935847549538e-05, "loss": 0.0496, "step": 12650 }, { "epoch": 0.2529874905079733, "grad_norm": 0.14569421112537384, "learning_rate": 9.976836595606543e-05, "loss": 0.046, "step": 12660 }, { "epoch": 0.25318732264897487, "grad_norm": 0.1453886181116104, "learning_rate": 9.97673713106223e-05, "loss": 0.0495, "step": 12670 }, { "epoch": 0.2533871547899764, "grad_norm": 0.11339522898197174, "learning_rate": 9.976637453920852e-05, "loss": 0.0459, "step": 12680 }, { "epoch": 0.253586986930978, "grad_norm": 0.13757577538490295, "learning_rate": 9.976537564186663e-05, "loss": 0.046, "step": 12690 }, { "epoch": 0.25378681907197953, "grad_norm": 0.16666002571582794, "learning_rate": 9.976437461863931e-05, "loss": 0.0423, "step": 12700 }, { "epoch": 0.2539866512129811, "grad_norm": 0.1543886661529541, "learning_rate": 9.976337146956934e-05, "loss": 0.0459, "step": 12710 }, { "epoch": 0.25418648335398264, "grad_norm": 0.18592101335525513, "learning_rate": 9.976236619469955e-05, "loss": 0.0553, "step": 12720 }, { "epoch": 0.2543863154949842, "grad_norm": 0.14417244493961334, "learning_rate": 9.976135879407288e-05, "loss": 0.0472, "step": 12730 }, { "epoch": 0.25458614763598575, "grad_norm": 0.1429569125175476, "learning_rate": 9.976034926773239e-05, "loss": 0.0466, "step": 12740 }, { "epoch": 0.2547859797769873, "grad_norm": 0.20693737268447876, "learning_rate": 9.975933761572119e-05, "loss": 0.05, "step": 12750 }, { "epoch": 0.2549858119179889, "grad_norm": 0.16053035855293274, "learning_rate": 9.97583238380825e-05, "loss": 0.0465, "step": 12760 }, { "epoch": 0.25518564405899047, "grad_norm": 0.18558260798454285, "learning_rate": 9.975730793485962e-05, "loss": 0.0557, "step": 12770 }, { "epoch": 0.255385476199992, "grad_norm": 0.16583117842674255, "learning_rate": 9.975628990609595e-05, "loss": 0.0497, "step": 12780 }, { "epoch": 0.2555853083409936, "grad_norm": 0.14223740994930267, "learning_rate": 9.9755269751835e-05, "loss": 0.0511, "step": 12790 }, { "epoch": 0.25578514048199513, "grad_norm": 0.165217325091362, "learning_rate": 9.975424747212032e-05, "loss": 0.0484, "step": 12800 }, { "epoch": 0.2559849726229967, "grad_norm": 0.15913964807987213, "learning_rate": 9.975322306699558e-05, "loss": 0.0543, "step": 12810 }, { "epoch": 0.25618480476399824, "grad_norm": 0.14956776797771454, "learning_rate": 9.975219653650456e-05, "loss": 0.0507, "step": 12820 }, { "epoch": 0.2563846369049998, "grad_norm": 0.14084629714488983, "learning_rate": 9.97511678806911e-05, "loss": 0.0448, "step": 12830 }, { "epoch": 0.25658446904600135, "grad_norm": 0.1441681981086731, "learning_rate": 9.975013709959914e-05, "loss": 0.0533, "step": 12840 }, { "epoch": 0.2567843011870029, "grad_norm": 0.16417689621448517, "learning_rate": 9.974910419327273e-05, "loss": 0.0473, "step": 12850 }, { "epoch": 0.25698413332800446, "grad_norm": 0.14471633732318878, "learning_rate": 9.974806916175596e-05, "loss": 0.0462, "step": 12860 }, { "epoch": 0.257183965469006, "grad_norm": 0.1616402566432953, "learning_rate": 9.974703200509308e-05, "loss": 0.0537, "step": 12870 }, { "epoch": 0.25738379761000757, "grad_norm": 0.1383787840604782, "learning_rate": 9.974599272332836e-05, "loss": 0.0484, "step": 12880 }, { "epoch": 0.2575836297510092, "grad_norm": 0.14418210089206696, "learning_rate": 9.974495131650624e-05, "loss": 0.0425, "step": 12890 }, { "epoch": 0.25778346189201073, "grad_norm": 0.18961162865161896, "learning_rate": 9.974390778467117e-05, "loss": 0.0482, "step": 12900 }, { "epoch": 0.2579832940330123, "grad_norm": 0.140452042222023, "learning_rate": 9.974286212786776e-05, "loss": 0.0409, "step": 12910 }, { "epoch": 0.25818312617401384, "grad_norm": 0.19099169969558716, "learning_rate": 9.974181434614064e-05, "loss": 0.0516, "step": 12920 }, { "epoch": 0.2583829583150154, "grad_norm": 0.18414519727230072, "learning_rate": 9.974076443953461e-05, "loss": 0.0486, "step": 12930 }, { "epoch": 0.25858279045601695, "grad_norm": 0.1642608642578125, "learning_rate": 9.973971240809449e-05, "loss": 0.0586, "step": 12940 }, { "epoch": 0.2587826225970185, "grad_norm": 0.19585998356342316, "learning_rate": 9.973865825186522e-05, "loss": 0.044, "step": 12950 }, { "epoch": 0.25898245473802006, "grad_norm": 0.14661453664302826, "learning_rate": 9.973760197089186e-05, "loss": 0.0499, "step": 12960 }, { "epoch": 0.2591822868790216, "grad_norm": 0.16108821332454681, "learning_rate": 9.973654356521951e-05, "loss": 0.0403, "step": 12970 }, { "epoch": 0.25938211902002317, "grad_norm": 0.1732034832239151, "learning_rate": 9.973548303489337e-05, "loss": 0.0522, "step": 12980 }, { "epoch": 0.2595819511610247, "grad_norm": 0.13121545314788818, "learning_rate": 9.97344203799588e-05, "loss": 0.0491, "step": 12990 }, { "epoch": 0.2597817833020263, "grad_norm": 0.2124571055173874, "learning_rate": 9.973335560046112e-05, "loss": 0.0483, "step": 13000 }, { "epoch": 0.25998161544302784, "grad_norm": 0.17303046584129333, "learning_rate": 9.973228869644586e-05, "loss": 0.0502, "step": 13010 }, { "epoch": 0.2601814475840294, "grad_norm": 0.1280394047498703, "learning_rate": 9.97312196679586e-05, "loss": 0.0517, "step": 13020 }, { "epoch": 0.260381279725031, "grad_norm": 0.20484791696071625, "learning_rate": 9.973014851504498e-05, "loss": 0.0477, "step": 13030 }, { "epoch": 0.26058111186603256, "grad_norm": 0.1843414306640625, "learning_rate": 9.97290752377508e-05, "loss": 0.0577, "step": 13040 }, { "epoch": 0.2607809440070341, "grad_norm": 0.1895289272069931, "learning_rate": 9.972799983612184e-05, "loss": 0.0604, "step": 13050 }, { "epoch": 0.26098077614803566, "grad_norm": 0.1569908708333969, "learning_rate": 9.972692231020411e-05, "loss": 0.0471, "step": 13060 }, { "epoch": 0.2611806082890372, "grad_norm": 0.1875704526901245, "learning_rate": 9.97258426600436e-05, "loss": 0.0469, "step": 13070 }, { "epoch": 0.2613804404300388, "grad_norm": 0.14536112546920776, "learning_rate": 9.972476088568644e-05, "loss": 0.0461, "step": 13080 }, { "epoch": 0.26158027257104033, "grad_norm": 0.14775486290454865, "learning_rate": 9.972367698717885e-05, "loss": 0.0484, "step": 13090 }, { "epoch": 0.2617801047120419, "grad_norm": 0.14167600870132446, "learning_rate": 9.972259096456712e-05, "loss": 0.0534, "step": 13100 }, { "epoch": 0.26197993685304344, "grad_norm": 0.2166980654001236, "learning_rate": 9.972150281789765e-05, "loss": 0.0565, "step": 13110 }, { "epoch": 0.262179768994045, "grad_norm": 0.17003066837787628, "learning_rate": 9.972041254721692e-05, "loss": 0.0499, "step": 13120 }, { "epoch": 0.26237960113504655, "grad_norm": 0.13473664224147797, "learning_rate": 9.97193201525715e-05, "loss": 0.0547, "step": 13130 }, { "epoch": 0.2625794332760481, "grad_norm": 0.23204174637794495, "learning_rate": 9.971822563400806e-05, "loss": 0.0533, "step": 13140 }, { "epoch": 0.26277926541704966, "grad_norm": 0.12831324338912964, "learning_rate": 9.971712899157334e-05, "loss": 0.064, "step": 13150 }, { "epoch": 0.2629790975580512, "grad_norm": 0.17371536791324615, "learning_rate": 9.971603022531424e-05, "loss": 0.0558, "step": 13160 }, { "epoch": 0.2631789296990528, "grad_norm": 0.1641884595155716, "learning_rate": 9.971492933527764e-05, "loss": 0.0488, "step": 13170 }, { "epoch": 0.2633787618400544, "grad_norm": 0.1789538860321045, "learning_rate": 9.971382632151059e-05, "loss": 0.0529, "step": 13180 }, { "epoch": 0.26357859398105593, "grad_norm": 0.15754324197769165, "learning_rate": 9.971272118406021e-05, "loss": 0.0463, "step": 13190 }, { "epoch": 0.2637784261220575, "grad_norm": 0.16792060434818268, "learning_rate": 9.97116139229737e-05, "loss": 0.0525, "step": 13200 }, { "epoch": 0.26397825826305904, "grad_norm": 0.14581915736198425, "learning_rate": 9.971050453829838e-05, "loss": 0.0459, "step": 13210 }, { "epoch": 0.2641780904040606, "grad_norm": 0.14868558943271637, "learning_rate": 9.970939303008161e-05, "loss": 0.056, "step": 13220 }, { "epoch": 0.26437792254506215, "grad_norm": 0.1775849461555481, "learning_rate": 9.97082793983709e-05, "loss": 0.0501, "step": 13230 }, { "epoch": 0.2645777546860637, "grad_norm": 0.1589037925004959, "learning_rate": 9.970716364321379e-05, "loss": 0.0446, "step": 13240 }, { "epoch": 0.26477758682706526, "grad_norm": 0.15268808603286743, "learning_rate": 9.9706045764658e-05, "loss": 0.0528, "step": 13250 }, { "epoch": 0.2649774189680668, "grad_norm": 0.17483584582805634, "learning_rate": 9.970492576275123e-05, "loss": 0.0554, "step": 13260 }, { "epoch": 0.26517725110906837, "grad_norm": 0.13636735081672668, "learning_rate": 9.970380363754134e-05, "loss": 0.0504, "step": 13270 }, { "epoch": 0.2653770832500699, "grad_norm": 0.22694967687129974, "learning_rate": 9.970267938907628e-05, "loss": 0.0524, "step": 13280 }, { "epoch": 0.2655769153910715, "grad_norm": 0.13963082432746887, "learning_rate": 9.970155301740403e-05, "loss": 0.0419, "step": 13290 }, { "epoch": 0.26577674753207303, "grad_norm": 0.14697253704071045, "learning_rate": 9.970042452257277e-05, "loss": 0.0494, "step": 13300 }, { "epoch": 0.26597657967307464, "grad_norm": 0.158907949924469, "learning_rate": 9.969929390463066e-05, "loss": 0.0476, "step": 13310 }, { "epoch": 0.2661764118140762, "grad_norm": 0.1425982266664505, "learning_rate": 9.969816116362603e-05, "loss": 0.0515, "step": 13320 }, { "epoch": 0.26637624395507775, "grad_norm": 0.4606921970844269, "learning_rate": 9.969702629960724e-05, "loss": 0.0503, "step": 13330 }, { "epoch": 0.2665760760960793, "grad_norm": 0.19626300036907196, "learning_rate": 9.96958893126228e-05, "loss": 0.0473, "step": 13340 }, { "epoch": 0.26677590823708086, "grad_norm": 0.1505875438451767, "learning_rate": 9.969475020272126e-05, "loss": 0.0594, "step": 13350 }, { "epoch": 0.2669757403780824, "grad_norm": 0.1142328754067421, "learning_rate": 9.969360896995127e-05, "loss": 0.0461, "step": 13360 }, { "epoch": 0.26717557251908397, "grad_norm": 0.162237286567688, "learning_rate": 9.96924656143616e-05, "loss": 0.043, "step": 13370 }, { "epoch": 0.2673754046600855, "grad_norm": 0.14097711443901062, "learning_rate": 9.969132013600109e-05, "loss": 0.0572, "step": 13380 }, { "epoch": 0.2675752368010871, "grad_norm": 0.13030186295509338, "learning_rate": 9.969017253491866e-05, "loss": 0.0465, "step": 13390 }, { "epoch": 0.26777506894208863, "grad_norm": 0.12990884482860565, "learning_rate": 9.968902281116335e-05, "loss": 0.0514, "step": 13400 }, { "epoch": 0.2679749010830902, "grad_norm": 0.17434248328208923, "learning_rate": 9.968787096478427e-05, "loss": 0.0446, "step": 13410 }, { "epoch": 0.26817473322409174, "grad_norm": 0.16522231698036194, "learning_rate": 9.968671699583062e-05, "loss": 0.054, "step": 13420 }, { "epoch": 0.2683745653650933, "grad_norm": 0.1646926999092102, "learning_rate": 9.968556090435171e-05, "loss": 0.0498, "step": 13430 }, { "epoch": 0.26857439750609485, "grad_norm": 0.19050076603889465, "learning_rate": 9.96844026903969e-05, "loss": 0.0489, "step": 13440 }, { "epoch": 0.26877422964709646, "grad_norm": 0.20798557996749878, "learning_rate": 9.968324235401567e-05, "loss": 0.0469, "step": 13450 }, { "epoch": 0.268974061788098, "grad_norm": 0.1620989739894867, "learning_rate": 9.968207989525762e-05, "loss": 0.0536, "step": 13460 }, { "epoch": 0.2691738939290996, "grad_norm": 0.18220913410186768, "learning_rate": 9.968091531417236e-05, "loss": 0.0607, "step": 13470 }, { "epoch": 0.2693737260701011, "grad_norm": 0.1410793960094452, "learning_rate": 9.967974861080969e-05, "loss": 0.0496, "step": 13480 }, { "epoch": 0.2695735582111027, "grad_norm": 0.14295637607574463, "learning_rate": 9.96785797852194e-05, "loss": 0.0488, "step": 13490 }, { "epoch": 0.26977339035210424, "grad_norm": 0.12529133260250092, "learning_rate": 9.967740883745147e-05, "loss": 0.0449, "step": 13500 }, { "epoch": 0.2699732224931058, "grad_norm": 0.17818984389305115, "learning_rate": 9.967623576755587e-05, "loss": 0.0417, "step": 13510 }, { "epoch": 0.27017305463410735, "grad_norm": 0.16468779742717743, "learning_rate": 9.967506057558275e-05, "loss": 0.05, "step": 13520 }, { "epoch": 0.2703728867751089, "grad_norm": 0.17865009605884552, "learning_rate": 9.96738832615823e-05, "loss": 0.0426, "step": 13530 }, { "epoch": 0.27057271891611046, "grad_norm": 0.20730464160442352, "learning_rate": 9.96727038256048e-05, "loss": 0.0459, "step": 13540 }, { "epoch": 0.270772551057112, "grad_norm": 0.15586668252944946, "learning_rate": 9.967152226770066e-05, "loss": 0.042, "step": 13550 }, { "epoch": 0.27097238319811356, "grad_norm": 0.15712812542915344, "learning_rate": 9.967033858792031e-05, "loss": 0.0518, "step": 13560 }, { "epoch": 0.2711722153391151, "grad_norm": 0.1815769076347351, "learning_rate": 9.966915278631437e-05, "loss": 0.0527, "step": 13570 }, { "epoch": 0.27137204748011673, "grad_norm": 0.13652844727039337, "learning_rate": 9.966796486293347e-05, "loss": 0.0466, "step": 13580 }, { "epoch": 0.2715718796211183, "grad_norm": 0.16633015871047974, "learning_rate": 9.966677481782835e-05, "loss": 0.0574, "step": 13590 }, { "epoch": 0.27177171176211984, "grad_norm": 0.16383770108222961, "learning_rate": 9.966558265104983e-05, "loss": 0.0609, "step": 13600 }, { "epoch": 0.2719715439031214, "grad_norm": 0.14791905879974365, "learning_rate": 9.966438836264888e-05, "loss": 0.0432, "step": 13610 }, { "epoch": 0.27217137604412295, "grad_norm": 0.13517220318317413, "learning_rate": 9.966319195267649e-05, "loss": 0.0444, "step": 13620 }, { "epoch": 0.2723712081851245, "grad_norm": 0.13973179459571838, "learning_rate": 9.966199342118377e-05, "loss": 0.0486, "step": 13630 }, { "epoch": 0.27257104032612606, "grad_norm": 0.1713007539510727, "learning_rate": 9.966079276822194e-05, "loss": 0.0429, "step": 13640 }, { "epoch": 0.2727708724671276, "grad_norm": 0.17230774462223053, "learning_rate": 9.965958999384224e-05, "loss": 0.0396, "step": 13650 }, { "epoch": 0.27297070460812917, "grad_norm": 0.14741139113903046, "learning_rate": 9.965838509809611e-05, "loss": 0.049, "step": 13660 }, { "epoch": 0.2731705367491307, "grad_norm": 0.14509809017181396, "learning_rate": 9.965717808103498e-05, "loss": 0.039, "step": 13670 }, { "epoch": 0.2733703688901323, "grad_norm": 0.14362548291683197, "learning_rate": 9.965596894271043e-05, "loss": 0.0529, "step": 13680 }, { "epoch": 0.27357020103113383, "grad_norm": 0.15697798132896423, "learning_rate": 9.965475768317411e-05, "loss": 0.0456, "step": 13690 }, { "epoch": 0.2737700331721354, "grad_norm": 0.1332056075334549, "learning_rate": 9.965354430247777e-05, "loss": 0.0425, "step": 13700 }, { "epoch": 0.27396986531313694, "grad_norm": 0.12943121790885925, "learning_rate": 9.965232880067321e-05, "loss": 0.0507, "step": 13710 }, { "epoch": 0.27416969745413855, "grad_norm": 0.1532508134841919, "learning_rate": 9.96511111778124e-05, "loss": 0.0546, "step": 13720 }, { "epoch": 0.2743695295951401, "grad_norm": 0.1457735300064087, "learning_rate": 9.964989143394732e-05, "loss": 0.0488, "step": 13730 }, { "epoch": 0.27456936173614166, "grad_norm": 0.4964990019798279, "learning_rate": 9.964866956913009e-05, "loss": 0.0574, "step": 13740 }, { "epoch": 0.2747691938771432, "grad_norm": 0.1727084219455719, "learning_rate": 9.96474455834129e-05, "loss": 0.0536, "step": 13750 }, { "epoch": 0.27496902601814477, "grad_norm": 0.2543964982032776, "learning_rate": 9.964621947684805e-05, "loss": 0.0592, "step": 13760 }, { "epoch": 0.2751688581591463, "grad_norm": 0.11831888556480408, "learning_rate": 9.96449912494879e-05, "loss": 0.0555, "step": 13770 }, { "epoch": 0.2753686903001479, "grad_norm": 0.13331736624240875, "learning_rate": 9.964376090138492e-05, "loss": 0.0519, "step": 13780 }, { "epoch": 0.27556852244114943, "grad_norm": 0.16769805550575256, "learning_rate": 9.964252843259168e-05, "loss": 0.0471, "step": 13790 }, { "epoch": 0.275768354582151, "grad_norm": 0.14343293011188507, "learning_rate": 9.964129384316083e-05, "loss": 0.0508, "step": 13800 }, { "epoch": 0.27596818672315254, "grad_norm": 0.2126842886209488, "learning_rate": 9.964005713314509e-05, "loss": 0.059, "step": 13810 }, { "epoch": 0.2761680188641541, "grad_norm": 0.14862626791000366, "learning_rate": 9.96388183025973e-05, "loss": 0.0453, "step": 13820 }, { "epoch": 0.27636785100515565, "grad_norm": 0.17618775367736816, "learning_rate": 9.963757735157038e-05, "loss": 0.0565, "step": 13830 }, { "epoch": 0.2765676831461572, "grad_norm": 0.18878942728042603, "learning_rate": 9.963633428011735e-05, "loss": 0.0573, "step": 13840 }, { "epoch": 0.27676751528715876, "grad_norm": 0.1951066255569458, "learning_rate": 9.96350890882913e-05, "loss": 0.0491, "step": 13850 }, { "epoch": 0.27696734742816037, "grad_norm": 0.12835706770420074, "learning_rate": 9.963384177614544e-05, "loss": 0.0423, "step": 13860 }, { "epoch": 0.2771671795691619, "grad_norm": 0.172257199883461, "learning_rate": 9.963259234373301e-05, "loss": 0.0445, "step": 13870 }, { "epoch": 0.2773670117101635, "grad_norm": 0.13749343156814575, "learning_rate": 9.963134079110743e-05, "loss": 0.0433, "step": 13880 }, { "epoch": 0.27756684385116503, "grad_norm": 0.15924741327762604, "learning_rate": 9.963008711832215e-05, "loss": 0.057, "step": 13890 }, { "epoch": 0.2777666759921666, "grad_norm": 0.11981333792209625, "learning_rate": 9.962883132543072e-05, "loss": 0.0549, "step": 13900 }, { "epoch": 0.27796650813316814, "grad_norm": 0.1693994253873825, "learning_rate": 9.962757341248679e-05, "loss": 0.0487, "step": 13910 }, { "epoch": 0.2781663402741697, "grad_norm": 0.14146830141544342, "learning_rate": 9.962631337954409e-05, "loss": 0.052, "step": 13920 }, { "epoch": 0.27836617241517125, "grad_norm": 0.14694377779960632, "learning_rate": 9.962505122665644e-05, "loss": 0.0415, "step": 13930 }, { "epoch": 0.2785660045561728, "grad_norm": 0.17712822556495667, "learning_rate": 9.962378695387777e-05, "loss": 0.0519, "step": 13940 }, { "epoch": 0.27876583669717436, "grad_norm": 0.14758244156837463, "learning_rate": 9.96225205612621e-05, "loss": 0.0461, "step": 13950 }, { "epoch": 0.2789656688381759, "grad_norm": 0.148576557636261, "learning_rate": 9.962125204886348e-05, "loss": 0.0494, "step": 13960 }, { "epoch": 0.27916550097917747, "grad_norm": 0.13338187336921692, "learning_rate": 9.961998141673615e-05, "loss": 0.0525, "step": 13970 }, { "epoch": 0.279365333120179, "grad_norm": 0.10291530936956406, "learning_rate": 9.961870866493437e-05, "loss": 0.051, "step": 13980 }, { "epoch": 0.2795651652611806, "grad_norm": 0.1505642533302307, "learning_rate": 9.961743379351251e-05, "loss": 0.0392, "step": 13990 }, { "epoch": 0.2797649974021822, "grad_norm": 0.17641553282737732, "learning_rate": 9.961615680252501e-05, "loss": 0.0553, "step": 14000 }, { "epoch": 0.27996482954318375, "grad_norm": 0.1745927780866623, "learning_rate": 9.961487769202646e-05, "loss": 0.0493, "step": 14010 }, { "epoch": 0.2801646616841853, "grad_norm": 0.14413519203662872, "learning_rate": 9.961359646207146e-05, "loss": 0.0513, "step": 14020 }, { "epoch": 0.28036449382518686, "grad_norm": 0.12824025750160217, "learning_rate": 9.961231311271479e-05, "loss": 0.045, "step": 14030 }, { "epoch": 0.2805643259661884, "grad_norm": 0.15415233373641968, "learning_rate": 9.961102764401122e-05, "loss": 0.0438, "step": 14040 }, { "epoch": 0.28076415810718997, "grad_norm": 0.15161530673503876, "learning_rate": 9.96097400560157e-05, "loss": 0.0496, "step": 14050 }, { "epoch": 0.2809639902481915, "grad_norm": 0.1539808064699173, "learning_rate": 9.960845034878321e-05, "loss": 0.0521, "step": 14060 }, { "epoch": 0.2811638223891931, "grad_norm": 0.1262030303478241, "learning_rate": 9.960715852236887e-05, "loss": 0.0464, "step": 14070 }, { "epoch": 0.28136365453019463, "grad_norm": 0.14801737666130066, "learning_rate": 9.960586457682783e-05, "loss": 0.0584, "step": 14080 }, { "epoch": 0.2815634866711962, "grad_norm": 0.1596699357032776, "learning_rate": 9.96045685122154e-05, "loss": 0.0476, "step": 14090 }, { "epoch": 0.28176331881219774, "grad_norm": 0.1471482813358307, "learning_rate": 9.960327032858692e-05, "loss": 0.057, "step": 14100 }, { "epoch": 0.2819631509531993, "grad_norm": 0.18534362316131592, "learning_rate": 9.960197002599786e-05, "loss": 0.045, "step": 14110 }, { "epoch": 0.28216298309420085, "grad_norm": 0.1499590426683426, "learning_rate": 9.960066760450378e-05, "loss": 0.0456, "step": 14120 }, { "epoch": 0.2823628152352024, "grad_norm": 0.203205406665802, "learning_rate": 9.959936306416027e-05, "loss": 0.0603, "step": 14130 }, { "epoch": 0.282562647376204, "grad_norm": 0.15521149337291718, "learning_rate": 9.959805640502311e-05, "loss": 0.0446, "step": 14140 }, { "epoch": 0.28276247951720557, "grad_norm": 0.165910542011261, "learning_rate": 9.959674762714808e-05, "loss": 0.0428, "step": 14150 }, { "epoch": 0.2829623116582071, "grad_norm": 0.13076519966125488, "learning_rate": 9.95954367305911e-05, "loss": 0.0457, "step": 14160 }, { "epoch": 0.2831621437992087, "grad_norm": 0.12669873237609863, "learning_rate": 9.959412371540818e-05, "loss": 0.0525, "step": 14170 }, { "epoch": 0.28336197594021023, "grad_norm": 0.16551271080970764, "learning_rate": 9.959280858165541e-05, "loss": 0.041, "step": 14180 }, { "epoch": 0.2835618080812118, "grad_norm": 0.18742337822914124, "learning_rate": 9.959149132938895e-05, "loss": 0.0517, "step": 14190 }, { "epoch": 0.28376164022221334, "grad_norm": 0.15170645713806152, "learning_rate": 9.95901719586651e-05, "loss": 0.0463, "step": 14200 }, { "epoch": 0.2839614723632149, "grad_norm": 0.14096033573150635, "learning_rate": 9.958885046954018e-05, "loss": 0.0509, "step": 14210 }, { "epoch": 0.28416130450421645, "grad_norm": 0.1424698680639267, "learning_rate": 9.958752686207067e-05, "loss": 0.0483, "step": 14220 }, { "epoch": 0.284361136645218, "grad_norm": 0.11463873833417892, "learning_rate": 9.958620113631312e-05, "loss": 0.0484, "step": 14230 }, { "epoch": 0.28456096878621956, "grad_norm": 0.18028560280799866, "learning_rate": 9.958487329232414e-05, "loss": 0.0569, "step": 14240 }, { "epoch": 0.2847608009272211, "grad_norm": 0.1429373323917389, "learning_rate": 9.958354333016047e-05, "loss": 0.0387, "step": 14250 }, { "epoch": 0.28496063306822267, "grad_norm": 0.12267760932445526, "learning_rate": 9.958221124987891e-05, "loss": 0.0484, "step": 14260 }, { "epoch": 0.2851604652092243, "grad_norm": 0.16923651099205017, "learning_rate": 9.958087705153638e-05, "loss": 0.0476, "step": 14270 }, { "epoch": 0.28536029735022583, "grad_norm": 0.17039084434509277, "learning_rate": 9.957954073518986e-05, "loss": 0.0545, "step": 14280 }, { "epoch": 0.2855601294912274, "grad_norm": 0.18223759531974792, "learning_rate": 9.957820230089645e-05, "loss": 0.0544, "step": 14290 }, { "epoch": 0.28575996163222894, "grad_norm": 0.13583318889141083, "learning_rate": 9.95768617487133e-05, "loss": 0.0441, "step": 14300 }, { "epoch": 0.2859597937732305, "grad_norm": 0.14968186616897583, "learning_rate": 9.95755190786977e-05, "loss": 0.0544, "step": 14310 }, { "epoch": 0.28615962591423205, "grad_norm": 0.14247263967990875, "learning_rate": 9.9574174290907e-05, "loss": 0.0473, "step": 14320 }, { "epoch": 0.2863594580552336, "grad_norm": 0.16740189492702484, "learning_rate": 9.957282738539864e-05, "loss": 0.0544, "step": 14330 }, { "epoch": 0.28655929019623516, "grad_norm": 0.15879036486148834, "learning_rate": 9.957147836223017e-05, "loss": 0.048, "step": 14340 }, { "epoch": 0.2867591223372367, "grad_norm": 0.10997170209884644, "learning_rate": 9.95701272214592e-05, "loss": 0.0477, "step": 14350 }, { "epoch": 0.28695895447823827, "grad_norm": 0.1707281768321991, "learning_rate": 9.956877396314346e-05, "loss": 0.0486, "step": 14360 }, { "epoch": 0.2871587866192398, "grad_norm": 0.23231147229671478, "learning_rate": 9.956741858734075e-05, "loss": 0.0528, "step": 14370 }, { "epoch": 0.2873586187602414, "grad_norm": 0.17904993891716003, "learning_rate": 9.9566061094109e-05, "loss": 0.055, "step": 14380 }, { "epoch": 0.28755845090124293, "grad_norm": 0.16489824652671814, "learning_rate": 9.956470148350617e-05, "loss": 0.0458, "step": 14390 }, { "epoch": 0.2877582830422445, "grad_norm": 0.11977013200521469, "learning_rate": 9.956333975559033e-05, "loss": 0.0484, "step": 14400 }, { "epoch": 0.2879581151832461, "grad_norm": 0.1620650440454483, "learning_rate": 9.956197591041967e-05, "loss": 0.0445, "step": 14410 }, { "epoch": 0.28815794732424765, "grad_norm": 0.15080048143863678, "learning_rate": 9.956060994805245e-05, "loss": 0.0407, "step": 14420 }, { "epoch": 0.2883577794652492, "grad_norm": 0.14055894315242767, "learning_rate": 9.955924186854701e-05, "loss": 0.0561, "step": 14430 }, { "epoch": 0.28855761160625076, "grad_norm": 0.1369810551404953, "learning_rate": 9.955787167196182e-05, "loss": 0.0531, "step": 14440 }, { "epoch": 0.2887574437472523, "grad_norm": 0.13252472877502441, "learning_rate": 9.955649935835539e-05, "loss": 0.0454, "step": 14450 }, { "epoch": 0.2889572758882539, "grad_norm": 0.18075238168239594, "learning_rate": 9.955512492778634e-05, "loss": 0.0475, "step": 14460 }, { "epoch": 0.2891571080292554, "grad_norm": 0.12907572090625763, "learning_rate": 9.955374838031338e-05, "loss": 0.0436, "step": 14470 }, { "epoch": 0.289356940170257, "grad_norm": 0.14917978644371033, "learning_rate": 9.955236971599533e-05, "loss": 0.0482, "step": 14480 }, { "epoch": 0.28955677231125854, "grad_norm": 0.14787402749061584, "learning_rate": 9.955098893489107e-05, "loss": 0.0458, "step": 14490 }, { "epoch": 0.2897566044522601, "grad_norm": 0.21764318645000458, "learning_rate": 9.954960603705961e-05, "loss": 0.046, "step": 14500 }, { "epoch": 0.28995643659326165, "grad_norm": 0.13724447786808014, "learning_rate": 9.954822102255998e-05, "loss": 0.0489, "step": 14510 }, { "epoch": 0.2901562687342632, "grad_norm": 0.18961045145988464, "learning_rate": 9.954683389145139e-05, "loss": 0.0495, "step": 14520 }, { "epoch": 0.29035610087526476, "grad_norm": 0.1866474598646164, "learning_rate": 9.954544464379307e-05, "loss": 0.0514, "step": 14530 }, { "epoch": 0.2905559330162663, "grad_norm": 0.12653927505016327, "learning_rate": 9.954405327964438e-05, "loss": 0.0486, "step": 14540 }, { "epoch": 0.2907557651572679, "grad_norm": 0.16308492422103882, "learning_rate": 9.954265979906474e-05, "loss": 0.0447, "step": 14550 }, { "epoch": 0.2909555972982695, "grad_norm": 0.13639280200004578, "learning_rate": 9.954126420211369e-05, "loss": 0.0432, "step": 14560 }, { "epoch": 0.29115542943927103, "grad_norm": 0.19123227894306183, "learning_rate": 9.953986648885084e-05, "loss": 0.0407, "step": 14570 }, { "epoch": 0.2913552615802726, "grad_norm": 0.159251868724823, "learning_rate": 9.953846665933591e-05, "loss": 0.0529, "step": 14580 }, { "epoch": 0.29155509372127414, "grad_norm": 0.15385770797729492, "learning_rate": 9.953706471362867e-05, "loss": 0.0599, "step": 14590 }, { "epoch": 0.2917549258622757, "grad_norm": 0.23029616475105286, "learning_rate": 9.953566065178905e-05, "loss": 0.0463, "step": 14600 }, { "epoch": 0.29195475800327725, "grad_norm": 0.13460038602352142, "learning_rate": 9.953425447387699e-05, "loss": 0.0485, "step": 14610 }, { "epoch": 0.2921545901442788, "grad_norm": 0.1818331927061081, "learning_rate": 9.953284617995259e-05, "loss": 0.0464, "step": 14620 }, { "epoch": 0.29235442228528036, "grad_norm": 0.1355290710926056, "learning_rate": 9.953143577007598e-05, "loss": 0.0461, "step": 14630 }, { "epoch": 0.2925542544262819, "grad_norm": 0.15869034826755524, "learning_rate": 9.953002324430744e-05, "loss": 0.0496, "step": 14640 }, { "epoch": 0.29275408656728347, "grad_norm": 0.14254847168922424, "learning_rate": 9.952860860270729e-05, "loss": 0.0502, "step": 14650 }, { "epoch": 0.292953918708285, "grad_norm": 0.13843053579330444, "learning_rate": 9.952719184533597e-05, "loss": 0.0508, "step": 14660 }, { "epoch": 0.2931537508492866, "grad_norm": 0.127624973654747, "learning_rate": 9.9525772972254e-05, "loss": 0.0442, "step": 14670 }, { "epoch": 0.29335358299028813, "grad_norm": 0.168064683675766, "learning_rate": 9.952435198352199e-05, "loss": 0.0514, "step": 14680 }, { "epoch": 0.29355341513128974, "grad_norm": 0.15065935254096985, "learning_rate": 9.952292887920065e-05, "loss": 0.0487, "step": 14690 }, { "epoch": 0.2937532472722913, "grad_norm": 0.16665531694889069, "learning_rate": 9.952150365935076e-05, "loss": 0.0433, "step": 14700 }, { "epoch": 0.29395307941329285, "grad_norm": 0.20696406066417694, "learning_rate": 9.952007632403319e-05, "loss": 0.0446, "step": 14710 }, { "epoch": 0.2941529115542944, "grad_norm": 0.13626419007778168, "learning_rate": 9.951864687330894e-05, "loss": 0.0456, "step": 14720 }, { "epoch": 0.29435274369529596, "grad_norm": 0.15444353222846985, "learning_rate": 9.951721530723908e-05, "loss": 0.0422, "step": 14730 }, { "epoch": 0.2945525758362975, "grad_norm": 0.14914058148860931, "learning_rate": 9.951578162588473e-05, "loss": 0.0514, "step": 14740 }, { "epoch": 0.29475240797729907, "grad_norm": 0.11988681554794312, "learning_rate": 9.951434582930716e-05, "loss": 0.0553, "step": 14750 }, { "epoch": 0.2949522401183006, "grad_norm": 0.17927739024162292, "learning_rate": 9.95129079175677e-05, "loss": 0.0458, "step": 14760 }, { "epoch": 0.2951520722593022, "grad_norm": 0.15054106712341309, "learning_rate": 9.951146789072777e-05, "loss": 0.0523, "step": 14770 }, { "epoch": 0.29535190440030373, "grad_norm": 0.13976065814495087, "learning_rate": 9.951002574884888e-05, "loss": 0.0445, "step": 14780 }, { "epoch": 0.2955517365413053, "grad_norm": 0.15252962708473206, "learning_rate": 9.950858149199264e-05, "loss": 0.0533, "step": 14790 }, { "epoch": 0.29575156868230684, "grad_norm": 0.10991097241640091, "learning_rate": 9.950713512022077e-05, "loss": 0.0454, "step": 14800 }, { "epoch": 0.2959514008233084, "grad_norm": 0.12570755183696747, "learning_rate": 9.950568663359502e-05, "loss": 0.0392, "step": 14810 }, { "epoch": 0.29615123296431, "grad_norm": 0.12907110154628754, "learning_rate": 9.950423603217728e-05, "loss": 0.0459, "step": 14820 }, { "epoch": 0.29635106510531156, "grad_norm": 0.15910035371780396, "learning_rate": 9.950278331602952e-05, "loss": 0.0514, "step": 14830 }, { "epoch": 0.2965508972463131, "grad_norm": 0.17069879174232483, "learning_rate": 9.950132848521381e-05, "loss": 0.0551, "step": 14840 }, { "epoch": 0.29675072938731467, "grad_norm": 0.13818353414535522, "learning_rate": 9.949987153979227e-05, "loss": 0.0443, "step": 14850 }, { "epoch": 0.2969505615283162, "grad_norm": 0.14885616302490234, "learning_rate": 9.949841247982714e-05, "loss": 0.0527, "step": 14860 }, { "epoch": 0.2971503936693178, "grad_norm": 0.1334080845117569, "learning_rate": 9.94969513053808e-05, "loss": 0.0504, "step": 14870 }, { "epoch": 0.29735022581031934, "grad_norm": 0.1674165576696396, "learning_rate": 9.949548801651558e-05, "loss": 0.0446, "step": 14880 }, { "epoch": 0.2975500579513209, "grad_norm": 0.23052532970905304, "learning_rate": 9.949402261329407e-05, "loss": 0.0523, "step": 14890 }, { "epoch": 0.29774989009232244, "grad_norm": 0.19938519597053528, "learning_rate": 9.949255509577883e-05, "loss": 0.0421, "step": 14900 }, { "epoch": 0.297949722233324, "grad_norm": 0.12579122185707092, "learning_rate": 9.949108546403257e-05, "loss": 0.0389, "step": 14910 }, { "epoch": 0.29814955437432555, "grad_norm": 0.1709429919719696, "learning_rate": 9.948961371811803e-05, "loss": 0.06, "step": 14920 }, { "epoch": 0.2983493865153271, "grad_norm": 0.17987409234046936, "learning_rate": 9.948813985809813e-05, "loss": 0.0508, "step": 14930 }, { "epoch": 0.29854921865632866, "grad_norm": 0.16980822384357452, "learning_rate": 9.948666388403579e-05, "loss": 0.0472, "step": 14940 }, { "epoch": 0.2987490507973302, "grad_norm": 0.14188137650489807, "learning_rate": 9.948518579599408e-05, "loss": 0.05, "step": 14950 }, { "epoch": 0.29894888293833183, "grad_norm": 0.18400311470031738, "learning_rate": 9.948370559403616e-05, "loss": 0.0471, "step": 14960 }, { "epoch": 0.2991487150793334, "grad_norm": 0.15232114493846893, "learning_rate": 9.948222327822523e-05, "loss": 0.0559, "step": 14970 }, { "epoch": 0.29934854722033494, "grad_norm": 0.1646413505077362, "learning_rate": 9.94807388486246e-05, "loss": 0.049, "step": 14980 }, { "epoch": 0.2995483793613365, "grad_norm": 0.19227111339569092, "learning_rate": 9.947925230529775e-05, "loss": 0.052, "step": 14990 }, { "epoch": 0.29974821150233805, "grad_norm": 0.16834446787834167, "learning_rate": 9.94777636483081e-05, "loss": 0.0463, "step": 15000 }, { "epoch": 0.2999480436433396, "grad_norm": 0.14616025984287262, "learning_rate": 9.94762728777193e-05, "loss": 0.0441, "step": 15010 }, { "epoch": 0.30014787578434116, "grad_norm": 0.15481112897396088, "learning_rate": 9.947477999359503e-05, "loss": 0.0411, "step": 15020 }, { "epoch": 0.3003477079253427, "grad_norm": 0.15160243213176727, "learning_rate": 9.947328499599901e-05, "loss": 0.051, "step": 15030 }, { "epoch": 0.30054754006634427, "grad_norm": 0.21629852056503296, "learning_rate": 9.947178788499516e-05, "loss": 0.0492, "step": 15040 }, { "epoch": 0.3007473722073458, "grad_norm": 0.16001969575881958, "learning_rate": 9.947028866064741e-05, "loss": 0.0449, "step": 15050 }, { "epoch": 0.3009472043483474, "grad_norm": 0.1831042617559433, "learning_rate": 9.946878732301981e-05, "loss": 0.0604, "step": 15060 }, { "epoch": 0.30114703648934893, "grad_norm": 0.15144452452659607, "learning_rate": 9.94672838721765e-05, "loss": 0.0477, "step": 15070 }, { "epoch": 0.3013468686303505, "grad_norm": 0.17797575891017914, "learning_rate": 9.946577830818168e-05, "loss": 0.0438, "step": 15080 }, { "epoch": 0.30154670077135204, "grad_norm": 0.20739613473415375, "learning_rate": 9.946427063109969e-05, "loss": 0.0477, "step": 15090 }, { "epoch": 0.30174653291235365, "grad_norm": 0.19574089348316193, "learning_rate": 9.946276084099493e-05, "loss": 0.0445, "step": 15100 }, { "epoch": 0.3019463650533552, "grad_norm": 0.12285454571247101, "learning_rate": 9.946124893793189e-05, "loss": 0.0434, "step": 15110 }, { "epoch": 0.30214619719435676, "grad_norm": 0.2503684163093567, "learning_rate": 9.945973492197516e-05, "loss": 0.0498, "step": 15120 }, { "epoch": 0.3023460293353583, "grad_norm": 0.14620035886764526, "learning_rate": 9.945821879318941e-05, "loss": 0.0382, "step": 15130 }, { "epoch": 0.30254586147635987, "grad_norm": 0.1666998565196991, "learning_rate": 9.945670055163942e-05, "loss": 0.054, "step": 15140 }, { "epoch": 0.3027456936173614, "grad_norm": 0.13103851675987244, "learning_rate": 9.945518019739004e-05, "loss": 0.0469, "step": 15150 }, { "epoch": 0.302945525758363, "grad_norm": 0.1465553641319275, "learning_rate": 9.94536577305062e-05, "loss": 0.0601, "step": 15160 }, { "epoch": 0.30314535789936453, "grad_norm": 0.1282104253768921, "learning_rate": 9.945213315105296e-05, "loss": 0.048, "step": 15170 }, { "epoch": 0.3033451900403661, "grad_norm": 0.14831435680389404, "learning_rate": 9.945060645909543e-05, "loss": 0.0493, "step": 15180 }, { "epoch": 0.30354502218136764, "grad_norm": 0.15933288633823395, "learning_rate": 9.944907765469884e-05, "loss": 0.0511, "step": 15190 }, { "epoch": 0.3037448543223692, "grad_norm": 0.14882956445217133, "learning_rate": 9.944754673792852e-05, "loss": 0.0533, "step": 15200 }, { "epoch": 0.30394468646337075, "grad_norm": 0.15727736055850983, "learning_rate": 9.94460137088498e-05, "loss": 0.041, "step": 15210 }, { "epoch": 0.3041445186043723, "grad_norm": 0.17164036631584167, "learning_rate": 9.944447856752824e-05, "loss": 0.048, "step": 15220 }, { "epoch": 0.30434435074537386, "grad_norm": 0.15879029035568237, "learning_rate": 9.944294131402936e-05, "loss": 0.0454, "step": 15230 }, { "epoch": 0.30454418288637547, "grad_norm": 0.15024423599243164, "learning_rate": 9.944140194841887e-05, "loss": 0.0433, "step": 15240 }, { "epoch": 0.304744015027377, "grad_norm": 0.15289510786533356, "learning_rate": 9.943986047076252e-05, "loss": 0.0494, "step": 15250 }, { "epoch": 0.3049438471683786, "grad_norm": 0.18844160437583923, "learning_rate": 9.943831688112615e-05, "loss": 0.0439, "step": 15260 }, { "epoch": 0.30514367930938013, "grad_norm": 0.12970715761184692, "learning_rate": 9.943677117957571e-05, "loss": 0.0522, "step": 15270 }, { "epoch": 0.3053435114503817, "grad_norm": 0.1580570638179779, "learning_rate": 9.943522336617721e-05, "loss": 0.0422, "step": 15280 }, { "epoch": 0.30554334359138324, "grad_norm": 0.16415689885616302, "learning_rate": 9.943367344099678e-05, "loss": 0.0522, "step": 15290 }, { "epoch": 0.3057431757323848, "grad_norm": 0.16527292132377625, "learning_rate": 9.943212140410066e-05, "loss": 0.0497, "step": 15300 }, { "epoch": 0.30594300787338635, "grad_norm": 0.1585799902677536, "learning_rate": 9.943056725555511e-05, "loss": 0.0502, "step": 15310 }, { "epoch": 0.3061428400143879, "grad_norm": 0.17981450259685516, "learning_rate": 9.942901099542653e-05, "loss": 0.0448, "step": 15320 }, { "epoch": 0.30634267215538946, "grad_norm": 0.1717849224805832, "learning_rate": 9.94274526237814e-05, "loss": 0.0507, "step": 15330 }, { "epoch": 0.306542504296391, "grad_norm": 0.12578639388084412, "learning_rate": 9.942589214068631e-05, "loss": 0.0451, "step": 15340 }, { "epoch": 0.30674233643739257, "grad_norm": 0.14831537008285522, "learning_rate": 9.942432954620788e-05, "loss": 0.0475, "step": 15350 }, { "epoch": 0.3069421685783941, "grad_norm": 0.15773864090442657, "learning_rate": 9.942276484041293e-05, "loss": 0.0454, "step": 15360 }, { "epoch": 0.3071420007193957, "grad_norm": 0.12392742186784744, "learning_rate": 9.942119802336822e-05, "loss": 0.0472, "step": 15370 }, { "epoch": 0.3073418328603973, "grad_norm": 0.16478247940540314, "learning_rate": 9.941962909514073e-05, "loss": 0.0418, "step": 15380 }, { "epoch": 0.30754166500139885, "grad_norm": 0.14255240559577942, "learning_rate": 9.941805805579746e-05, "loss": 0.0466, "step": 15390 }, { "epoch": 0.3077414971424004, "grad_norm": 0.1862839013338089, "learning_rate": 9.941648490540555e-05, "loss": 0.0488, "step": 15400 }, { "epoch": 0.30794132928340195, "grad_norm": 0.16222429275512695, "learning_rate": 9.94149096440322e-05, "loss": 0.0473, "step": 15410 }, { "epoch": 0.3081411614244035, "grad_norm": 0.16772381961345673, "learning_rate": 9.941333227174467e-05, "loss": 0.0538, "step": 15420 }, { "epoch": 0.30834099356540506, "grad_norm": 0.18345069885253906, "learning_rate": 9.941175278861037e-05, "loss": 0.045, "step": 15430 }, { "epoch": 0.3085408257064066, "grad_norm": 0.20834475755691528, "learning_rate": 9.941017119469675e-05, "loss": 0.0469, "step": 15440 }, { "epoch": 0.3087406578474082, "grad_norm": 0.1403486579656601, "learning_rate": 9.94085874900714e-05, "loss": 0.0518, "step": 15450 }, { "epoch": 0.3089404899884097, "grad_norm": 0.15243256092071533, "learning_rate": 9.940700167480194e-05, "loss": 0.0555, "step": 15460 }, { "epoch": 0.3091403221294113, "grad_norm": 0.15276023745536804, "learning_rate": 9.940541374895615e-05, "loss": 0.0395, "step": 15470 }, { "epoch": 0.30934015427041284, "grad_norm": 0.11902028322219849, "learning_rate": 9.940382371260184e-05, "loss": 0.0461, "step": 15480 }, { "epoch": 0.3095399864114144, "grad_norm": 0.13964593410491943, "learning_rate": 9.940223156580693e-05, "loss": 0.0449, "step": 15490 }, { "epoch": 0.30973981855241595, "grad_norm": 0.16042211651802063, "learning_rate": 9.940063730863945e-05, "loss": 0.0511, "step": 15500 }, { "epoch": 0.30993965069341756, "grad_norm": 0.18531255424022675, "learning_rate": 9.93990409411675e-05, "loss": 0.0414, "step": 15510 }, { "epoch": 0.3101394828344191, "grad_norm": 0.1256147176027298, "learning_rate": 9.939744246345926e-05, "loss": 0.0498, "step": 15520 }, { "epoch": 0.31033931497542067, "grad_norm": 0.12690843641757965, "learning_rate": 9.939584187558303e-05, "loss": 0.0385, "step": 15530 }, { "epoch": 0.3105391471164222, "grad_norm": 0.15873657166957855, "learning_rate": 9.939423917760717e-05, "loss": 0.044, "step": 15540 }, { "epoch": 0.3107389792574238, "grad_norm": 0.12983281910419464, "learning_rate": 9.939263436960017e-05, "loss": 0.0495, "step": 15550 }, { "epoch": 0.31093881139842533, "grad_norm": 0.18264296650886536, "learning_rate": 9.939102745163055e-05, "loss": 0.0493, "step": 15560 }, { "epoch": 0.3111386435394269, "grad_norm": 0.13941709697246552, "learning_rate": 9.9389418423767e-05, "loss": 0.0469, "step": 15570 }, { "epoch": 0.31133847568042844, "grad_norm": 0.16528815031051636, "learning_rate": 9.93878072860782e-05, "loss": 0.0427, "step": 15580 }, { "epoch": 0.31153830782143, "grad_norm": 0.13293428719043732, "learning_rate": 9.938619403863298e-05, "loss": 0.054, "step": 15590 }, { "epoch": 0.31173813996243155, "grad_norm": 0.20398911833763123, "learning_rate": 9.938457868150031e-05, "loss": 0.0381, "step": 15600 }, { "epoch": 0.3119379721034331, "grad_norm": 0.12645524740219116, "learning_rate": 9.938296121474916e-05, "loss": 0.0446, "step": 15610 }, { "epoch": 0.31213780424443466, "grad_norm": 0.19003409147262573, "learning_rate": 9.938134163844861e-05, "loss": 0.0527, "step": 15620 }, { "epoch": 0.3123376363854362, "grad_norm": 0.1393168568611145, "learning_rate": 9.937971995266789e-05, "loss": 0.0575, "step": 15630 }, { "epoch": 0.31253746852643777, "grad_norm": 0.15504638850688934, "learning_rate": 9.937809615747622e-05, "loss": 0.0455, "step": 15640 }, { "epoch": 0.3127373006674394, "grad_norm": 0.14236293733119965, "learning_rate": 9.937647025294298e-05, "loss": 0.0581, "step": 15650 }, { "epoch": 0.31293713280844093, "grad_norm": 0.16567564010620117, "learning_rate": 9.937484223913767e-05, "loss": 0.0544, "step": 15660 }, { "epoch": 0.3131369649494425, "grad_norm": 0.14339405298233032, "learning_rate": 9.93732121161298e-05, "loss": 0.0453, "step": 15670 }, { "epoch": 0.31333679709044404, "grad_norm": 0.15298718214035034, "learning_rate": 9.937157988398899e-05, "loss": 0.0512, "step": 15680 }, { "epoch": 0.3135366292314456, "grad_norm": 0.160856693983078, "learning_rate": 9.936994554278501e-05, "loss": 0.0486, "step": 15690 }, { "epoch": 0.31373646137244715, "grad_norm": 0.16650190949440002, "learning_rate": 9.936830909258763e-05, "loss": 0.0449, "step": 15700 }, { "epoch": 0.3139362935134487, "grad_norm": 0.17832082509994507, "learning_rate": 9.93666705334668e-05, "loss": 0.0414, "step": 15710 }, { "epoch": 0.31413612565445026, "grad_norm": 0.16177956759929657, "learning_rate": 9.936502986549248e-05, "loss": 0.0454, "step": 15720 }, { "epoch": 0.3143359577954518, "grad_norm": 0.16788265109062195, "learning_rate": 9.936338708873477e-05, "loss": 0.0487, "step": 15730 }, { "epoch": 0.31453578993645337, "grad_norm": 0.19729800522327423, "learning_rate": 9.936174220326385e-05, "loss": 0.0431, "step": 15740 }, { "epoch": 0.3147356220774549, "grad_norm": 0.1523110568523407, "learning_rate": 9.936009520914998e-05, "loss": 0.0534, "step": 15750 }, { "epoch": 0.3149354542184565, "grad_norm": 0.23619253933429718, "learning_rate": 9.935844610646352e-05, "loss": 0.0473, "step": 15760 }, { "epoch": 0.31513528635945803, "grad_norm": 0.15290041267871857, "learning_rate": 9.935679489527493e-05, "loss": 0.0515, "step": 15770 }, { "epoch": 0.3153351185004596, "grad_norm": 0.16309842467308044, "learning_rate": 9.935514157565472e-05, "loss": 0.0479, "step": 15780 }, { "epoch": 0.3155349506414612, "grad_norm": 0.20889535546302795, "learning_rate": 9.935348614767353e-05, "loss": 0.0484, "step": 15790 }, { "epoch": 0.31573478278246275, "grad_norm": 0.12827444076538086, "learning_rate": 9.935182861140209e-05, "loss": 0.0413, "step": 15800 }, { "epoch": 0.3159346149234643, "grad_norm": 0.15695929527282715, "learning_rate": 9.935016896691118e-05, "loss": 0.0587, "step": 15810 }, { "epoch": 0.31613444706446586, "grad_norm": 0.13505525887012482, "learning_rate": 9.934850721427171e-05, "loss": 0.0512, "step": 15820 }, { "epoch": 0.3163342792054674, "grad_norm": 0.14996550977230072, "learning_rate": 9.934684335355468e-05, "loss": 0.05, "step": 15830 }, { "epoch": 0.31653411134646897, "grad_norm": 0.17440107464790344, "learning_rate": 9.934517738483115e-05, "loss": 0.0469, "step": 15840 }, { "epoch": 0.3167339434874705, "grad_norm": 0.16357052326202393, "learning_rate": 9.934350930817228e-05, "loss": 0.0434, "step": 15850 }, { "epoch": 0.3169337756284721, "grad_norm": 0.1769929677248001, "learning_rate": 9.934183912364936e-05, "loss": 0.0575, "step": 15860 }, { "epoch": 0.31713360776947364, "grad_norm": 0.3953201472759247, "learning_rate": 9.934016683133371e-05, "loss": 0.0477, "step": 15870 }, { "epoch": 0.3173334399104752, "grad_norm": 0.1917683333158493, "learning_rate": 9.933849243129677e-05, "loss": 0.0513, "step": 15880 }, { "epoch": 0.31753327205147674, "grad_norm": 0.16488389670848846, "learning_rate": 9.933681592361006e-05, "loss": 0.0573, "step": 15890 }, { "epoch": 0.3177331041924783, "grad_norm": 0.14278966188430786, "learning_rate": 9.933513730834523e-05, "loss": 0.0536, "step": 15900 }, { "epoch": 0.31793293633347985, "grad_norm": 0.16109149158000946, "learning_rate": 9.933345658557396e-05, "loss": 0.054, "step": 15910 }, { "epoch": 0.3181327684744814, "grad_norm": 0.1376589983701706, "learning_rate": 9.933177375536803e-05, "loss": 0.0584, "step": 15920 }, { "epoch": 0.318332600615483, "grad_norm": 0.15466046333312988, "learning_rate": 9.933008881779937e-05, "loss": 0.0502, "step": 15930 }, { "epoch": 0.3185324327564846, "grad_norm": 0.1463472992181778, "learning_rate": 9.932840177293994e-05, "loss": 0.0527, "step": 15940 }, { "epoch": 0.31873226489748613, "grad_norm": 0.1568177044391632, "learning_rate": 9.932671262086182e-05, "loss": 0.0509, "step": 15950 }, { "epoch": 0.3189320970384877, "grad_norm": 0.1440296769142151, "learning_rate": 9.932502136163712e-05, "loss": 0.0569, "step": 15960 }, { "epoch": 0.31913192917948924, "grad_norm": 0.13569961488246918, "learning_rate": 9.932332799533815e-05, "loss": 0.0432, "step": 15970 }, { "epoch": 0.3193317613204908, "grad_norm": 0.1500367373228073, "learning_rate": 9.93216325220372e-05, "loss": 0.0535, "step": 15980 }, { "epoch": 0.31953159346149235, "grad_norm": 0.15953338146209717, "learning_rate": 9.931993494180672e-05, "loss": 0.0445, "step": 15990 }, { "epoch": 0.3197314256024939, "grad_norm": 0.17757108807563782, "learning_rate": 9.931823525471922e-05, "loss": 0.052, "step": 16000 }, { "epoch": 0.31993125774349546, "grad_norm": 0.1980510950088501, "learning_rate": 9.931653346084731e-05, "loss": 0.0539, "step": 16010 }, { "epoch": 0.320131089884497, "grad_norm": 0.15986168384552002, "learning_rate": 9.93148295602637e-05, "loss": 0.0535, "step": 16020 }, { "epoch": 0.32033092202549857, "grad_norm": 0.2140287309885025, "learning_rate": 9.931312355304115e-05, "loss": 0.055, "step": 16030 }, { "epoch": 0.3205307541665001, "grad_norm": 0.17340686917304993, "learning_rate": 9.931141543925255e-05, "loss": 0.0439, "step": 16040 }, { "epoch": 0.3207305863075017, "grad_norm": 0.20759472250938416, "learning_rate": 9.930970521897089e-05, "loss": 0.0524, "step": 16050 }, { "epoch": 0.32093041844850323, "grad_norm": 0.15572454035282135, "learning_rate": 9.930799289226919e-05, "loss": 0.0469, "step": 16060 }, { "epoch": 0.32113025058950484, "grad_norm": 0.1942446231842041, "learning_rate": 9.930627845922062e-05, "loss": 0.0492, "step": 16070 }, { "epoch": 0.3213300827305064, "grad_norm": 0.16226527094841003, "learning_rate": 9.930456191989841e-05, "loss": 0.0393, "step": 16080 }, { "epoch": 0.32152991487150795, "grad_norm": 0.14253677427768707, "learning_rate": 9.93028432743759e-05, "loss": 0.0447, "step": 16090 }, { "epoch": 0.3217297470125095, "grad_norm": 0.1916462928056717, "learning_rate": 9.930112252272649e-05, "loss": 0.0412, "step": 16100 }, { "epoch": 0.32192957915351106, "grad_norm": 0.15899226069450378, "learning_rate": 9.92993996650237e-05, "loss": 0.0519, "step": 16110 }, { "epoch": 0.3221294112945126, "grad_norm": 0.12464641034603119, "learning_rate": 9.929767470134112e-05, "loss": 0.0479, "step": 16120 }, { "epoch": 0.32232924343551417, "grad_norm": 0.14276333153247833, "learning_rate": 9.929594763175244e-05, "loss": 0.0523, "step": 16130 }, { "epoch": 0.3225290755765157, "grad_norm": 0.13097086548805237, "learning_rate": 9.929421845633144e-05, "loss": 0.0393, "step": 16140 }, { "epoch": 0.3227289077175173, "grad_norm": 0.1595982313156128, "learning_rate": 9.929248717515198e-05, "loss": 0.0456, "step": 16150 }, { "epoch": 0.32292873985851883, "grad_norm": 0.17044487595558167, "learning_rate": 9.929075378828802e-05, "loss": 0.0463, "step": 16160 }, { "epoch": 0.3231285719995204, "grad_norm": 0.1457056850194931, "learning_rate": 9.928901829581361e-05, "loss": 0.0458, "step": 16170 }, { "epoch": 0.32332840414052194, "grad_norm": 0.14657124876976013, "learning_rate": 9.928728069780289e-05, "loss": 0.0434, "step": 16180 }, { "epoch": 0.3235282362815235, "grad_norm": 0.1842557042837143, "learning_rate": 9.928554099433007e-05, "loss": 0.0458, "step": 16190 }, { "epoch": 0.3237280684225251, "grad_norm": 0.13698354363441467, "learning_rate": 9.928379918546949e-05, "loss": 0.0401, "step": 16200 }, { "epoch": 0.32392790056352666, "grad_norm": 0.1731492578983307, "learning_rate": 9.928205527129553e-05, "loss": 0.0459, "step": 16210 }, { "epoch": 0.3241277327045282, "grad_norm": 0.18127509951591492, "learning_rate": 9.928030925188271e-05, "loss": 0.0453, "step": 16220 }, { "epoch": 0.32432756484552977, "grad_norm": 0.16358928382396698, "learning_rate": 9.927856112730563e-05, "loss": 0.0448, "step": 16230 }, { "epoch": 0.3245273969865313, "grad_norm": 0.13873247802257538, "learning_rate": 9.927681089763892e-05, "loss": 0.0538, "step": 16240 }, { "epoch": 0.3247272291275329, "grad_norm": 0.14374098181724548, "learning_rate": 9.927505856295738e-05, "loss": 0.0441, "step": 16250 }, { "epoch": 0.32492706126853443, "grad_norm": 0.14172516763210297, "learning_rate": 9.927330412333586e-05, "loss": 0.04, "step": 16260 }, { "epoch": 0.325126893409536, "grad_norm": 0.13833175599575043, "learning_rate": 9.927154757884931e-05, "loss": 0.0421, "step": 16270 }, { "epoch": 0.32532672555053754, "grad_norm": 0.12610313296318054, "learning_rate": 9.926978892957277e-05, "loss": 0.0574, "step": 16280 }, { "epoch": 0.3255265576915391, "grad_norm": 0.1485769897699356, "learning_rate": 9.926802817558135e-05, "loss": 0.0496, "step": 16290 }, { "epoch": 0.32572638983254065, "grad_norm": 0.11576549708843231, "learning_rate": 9.926626531695028e-05, "loss": 0.049, "step": 16300 }, { "epoch": 0.3259262219735422, "grad_norm": 0.14501240849494934, "learning_rate": 9.926450035375484e-05, "loss": 0.0514, "step": 16310 }, { "epoch": 0.32612605411454376, "grad_norm": 0.1680220067501068, "learning_rate": 9.926273328607047e-05, "loss": 0.0428, "step": 16320 }, { "epoch": 0.3263258862555453, "grad_norm": 0.14692765474319458, "learning_rate": 9.926096411397261e-05, "loss": 0.0436, "step": 16330 }, { "epoch": 0.3265257183965469, "grad_norm": 0.15319427847862244, "learning_rate": 9.925919283753688e-05, "loss": 0.0471, "step": 16340 }, { "epoch": 0.3267255505375485, "grad_norm": 0.16964854300022125, "learning_rate": 9.925741945683891e-05, "loss": 0.053, "step": 16350 }, { "epoch": 0.32692538267855004, "grad_norm": 0.1612362265586853, "learning_rate": 9.925564397195448e-05, "loss": 0.0454, "step": 16360 }, { "epoch": 0.3271252148195516, "grad_norm": 0.1687595099210739, "learning_rate": 9.925386638295941e-05, "loss": 0.0507, "step": 16370 }, { "epoch": 0.32732504696055315, "grad_norm": 0.12852217257022858, "learning_rate": 9.925208668992967e-05, "loss": 0.0463, "step": 16380 }, { "epoch": 0.3275248791015547, "grad_norm": 0.14601507782936096, "learning_rate": 9.925030489294126e-05, "loss": 0.0567, "step": 16390 }, { "epoch": 0.32772471124255625, "grad_norm": 0.18808597326278687, "learning_rate": 9.92485209920703e-05, "loss": 0.0415, "step": 16400 }, { "epoch": 0.3279245433835578, "grad_norm": 0.18002401292324066, "learning_rate": 9.924673498739299e-05, "loss": 0.0472, "step": 16410 }, { "epoch": 0.32812437552455936, "grad_norm": 0.16679970920085907, "learning_rate": 9.924494687898563e-05, "loss": 0.0429, "step": 16420 }, { "epoch": 0.3283242076655609, "grad_norm": 0.14990735054016113, "learning_rate": 9.92431566669246e-05, "loss": 0.0561, "step": 16430 }, { "epoch": 0.3285240398065625, "grad_norm": 0.16291667520999908, "learning_rate": 9.924136435128639e-05, "loss": 0.0456, "step": 16440 }, { "epoch": 0.32872387194756403, "grad_norm": 0.21368375420570374, "learning_rate": 9.923956993214755e-05, "loss": 0.0471, "step": 16450 }, { "epoch": 0.3289237040885656, "grad_norm": 0.15883998572826385, "learning_rate": 9.923777340958476e-05, "loss": 0.0544, "step": 16460 }, { "epoch": 0.32912353622956714, "grad_norm": 0.15330876410007477, "learning_rate": 9.923597478367471e-05, "loss": 0.0413, "step": 16470 }, { "epoch": 0.32932336837056875, "grad_norm": 0.12933549284934998, "learning_rate": 9.923417405449429e-05, "loss": 0.0471, "step": 16480 }, { "epoch": 0.3295232005115703, "grad_norm": 0.1427530199289322, "learning_rate": 9.923237122212039e-05, "loss": 0.0422, "step": 16490 }, { "epoch": 0.32972303265257186, "grad_norm": 0.1278664767742157, "learning_rate": 9.923056628663004e-05, "loss": 0.0423, "step": 16500 }, { "epoch": 0.3299228647935734, "grad_norm": 0.151303231716156, "learning_rate": 9.922875924810032e-05, "loss": 0.0453, "step": 16510 }, { "epoch": 0.33012269693457497, "grad_norm": 0.19310088455677032, "learning_rate": 9.922695010660845e-05, "loss": 0.0456, "step": 16520 }, { "epoch": 0.3303225290755765, "grad_norm": 0.16829673945903778, "learning_rate": 9.922513886223172e-05, "loss": 0.0608, "step": 16530 }, { "epoch": 0.3305223612165781, "grad_norm": 0.21674218773841858, "learning_rate": 9.922332551504747e-05, "loss": 0.0568, "step": 16540 }, { "epoch": 0.33072219335757963, "grad_norm": 0.11419602483510971, "learning_rate": 9.922151006513319e-05, "loss": 0.0451, "step": 16550 }, { "epoch": 0.3309220254985812, "grad_norm": 0.13291800022125244, "learning_rate": 9.921969251256642e-05, "loss": 0.0428, "step": 16560 }, { "epoch": 0.33112185763958274, "grad_norm": 0.14368893206119537, "learning_rate": 9.921787285742479e-05, "loss": 0.0561, "step": 16570 }, { "epoch": 0.3313216897805843, "grad_norm": 0.15501825511455536, "learning_rate": 9.921605109978607e-05, "loss": 0.0411, "step": 16580 }, { "epoch": 0.33152152192158585, "grad_norm": 0.1641431748867035, "learning_rate": 9.921422723972804e-05, "loss": 0.0464, "step": 16590 }, { "epoch": 0.3317213540625874, "grad_norm": 0.15544500946998596, "learning_rate": 9.921240127732865e-05, "loss": 0.0358, "step": 16600 }, { "epoch": 0.33192118620358896, "grad_norm": 0.11925584822893143, "learning_rate": 9.921057321266587e-05, "loss": 0.0478, "step": 16610 }, { "epoch": 0.33212101834459057, "grad_norm": 0.1379024237394333, "learning_rate": 9.920874304581782e-05, "loss": 0.0526, "step": 16620 }, { "epoch": 0.3323208504855921, "grad_norm": 0.19287878274917603, "learning_rate": 9.920691077686265e-05, "loss": 0.0509, "step": 16630 }, { "epoch": 0.3325206826265937, "grad_norm": 0.15451344847679138, "learning_rate": 9.920507640587866e-05, "loss": 0.0514, "step": 16640 }, { "epoch": 0.33272051476759523, "grad_norm": 0.14973653852939606, "learning_rate": 9.920323993294419e-05, "loss": 0.0444, "step": 16650 }, { "epoch": 0.3329203469085968, "grad_norm": 0.11167571693658829, "learning_rate": 9.92014013581377e-05, "loss": 0.0507, "step": 16660 }, { "epoch": 0.33312017904959834, "grad_norm": 0.18194453418254852, "learning_rate": 9.919956068153773e-05, "loss": 0.0466, "step": 16670 }, { "epoch": 0.3333200111905999, "grad_norm": 0.16851812601089478, "learning_rate": 9.919771790322292e-05, "loss": 0.0423, "step": 16680 }, { "epoch": 0.33351984333160145, "grad_norm": 0.17212992906570435, "learning_rate": 9.919587302327198e-05, "loss": 0.0551, "step": 16690 }, { "epoch": 0.333719675472603, "grad_norm": 0.15690955519676208, "learning_rate": 9.919402604176372e-05, "loss": 0.0459, "step": 16700 }, { "epoch": 0.33391950761360456, "grad_norm": 0.17556016147136688, "learning_rate": 9.919217695877701e-05, "loss": 0.0477, "step": 16710 }, { "epoch": 0.3341193397546061, "grad_norm": 0.12262386828660965, "learning_rate": 9.91903257743909e-05, "loss": 0.0488, "step": 16720 }, { "epoch": 0.33431917189560767, "grad_norm": 0.13321644067764282, "learning_rate": 9.918847248868443e-05, "loss": 0.0385, "step": 16730 }, { "epoch": 0.3345190040366092, "grad_norm": 0.13635797798633575, "learning_rate": 9.918661710173678e-05, "loss": 0.0411, "step": 16740 }, { "epoch": 0.33471883617761083, "grad_norm": 0.13469639420509338, "learning_rate": 9.918475961362721e-05, "loss": 0.0453, "step": 16750 }, { "epoch": 0.3349186683186124, "grad_norm": 0.143583282828331, "learning_rate": 9.918290002443504e-05, "loss": 0.0491, "step": 16760 }, { "epoch": 0.33511850045961394, "grad_norm": 0.18214760720729828, "learning_rate": 9.918103833423977e-05, "loss": 0.0474, "step": 16770 }, { "epoch": 0.3353183326006155, "grad_norm": 0.12009401619434357, "learning_rate": 9.917917454312086e-05, "loss": 0.049, "step": 16780 }, { "epoch": 0.33551816474161705, "grad_norm": 0.12710000574588776, "learning_rate": 9.917730865115798e-05, "loss": 0.0447, "step": 16790 }, { "epoch": 0.3357179968826186, "grad_norm": 0.1252213716506958, "learning_rate": 9.91754406584308e-05, "loss": 0.0372, "step": 16800 }, { "epoch": 0.33591782902362016, "grad_norm": 0.14596199989318848, "learning_rate": 9.917357056501915e-05, "loss": 0.0442, "step": 16810 }, { "epoch": 0.3361176611646217, "grad_norm": 0.12605050206184387, "learning_rate": 9.917169837100289e-05, "loss": 0.0486, "step": 16820 }, { "epoch": 0.33631749330562327, "grad_norm": 0.18953333795070648, "learning_rate": 9.9169824076462e-05, "loss": 0.0435, "step": 16830 }, { "epoch": 0.3365173254466248, "grad_norm": 0.1559208631515503, "learning_rate": 9.916794768147657e-05, "loss": 0.0444, "step": 16840 }, { "epoch": 0.3367171575876264, "grad_norm": 0.11009345948696136, "learning_rate": 9.916606918612672e-05, "loss": 0.0423, "step": 16850 }, { "epoch": 0.33691698972862794, "grad_norm": 0.1606566458940506, "learning_rate": 9.916418859049272e-05, "loss": 0.0438, "step": 16860 }, { "epoch": 0.3371168218696295, "grad_norm": 0.19388695061206818, "learning_rate": 9.916230589465491e-05, "loss": 0.0401, "step": 16870 }, { "epoch": 0.33731665401063105, "grad_norm": 0.159465491771698, "learning_rate": 9.91604210986937e-05, "loss": 0.045, "step": 16880 }, { "epoch": 0.33751648615163266, "grad_norm": 0.14265775680541992, "learning_rate": 9.915853420268963e-05, "loss": 0.0446, "step": 16890 }, { "epoch": 0.3377163182926342, "grad_norm": 0.12091207504272461, "learning_rate": 9.915664520672325e-05, "loss": 0.0474, "step": 16900 }, { "epoch": 0.33791615043363576, "grad_norm": 0.12293339520692825, "learning_rate": 9.915475411087531e-05, "loss": 0.0392, "step": 16910 }, { "epoch": 0.3381159825746373, "grad_norm": 0.15531815588474274, "learning_rate": 9.915286091522659e-05, "loss": 0.0427, "step": 16920 }, { "epoch": 0.3383158147156389, "grad_norm": 0.1678028404712677, "learning_rate": 9.915096561985793e-05, "loss": 0.0461, "step": 16930 }, { "epoch": 0.33851564685664043, "grad_norm": 0.18068362772464752, "learning_rate": 9.914906822485031e-05, "loss": 0.0433, "step": 16940 }, { "epoch": 0.338715478997642, "grad_norm": 0.1515423059463501, "learning_rate": 9.91471687302848e-05, "loss": 0.0419, "step": 16950 }, { "epoch": 0.33891531113864354, "grad_norm": 0.185137078166008, "learning_rate": 9.914526713624252e-05, "loss": 0.0616, "step": 16960 }, { "epoch": 0.3391151432796451, "grad_norm": 0.14011146128177643, "learning_rate": 9.914336344280471e-05, "loss": 0.0397, "step": 16970 }, { "epoch": 0.33931497542064665, "grad_norm": 0.19188635051250458, "learning_rate": 9.914145765005268e-05, "loss": 0.0558, "step": 16980 }, { "epoch": 0.3395148075616482, "grad_norm": 0.16990496218204498, "learning_rate": 9.913954975806787e-05, "loss": 0.0411, "step": 16990 }, { "epoch": 0.33971463970264976, "grad_norm": 0.17576774954795837, "learning_rate": 9.913763976693177e-05, "loss": 0.0458, "step": 17000 }, { "epoch": 0.3399144718436513, "grad_norm": 0.1270146369934082, "learning_rate": 9.913572767672597e-05, "loss": 0.0475, "step": 17010 }, { "epoch": 0.34011430398465287, "grad_norm": 0.12989607453346252, "learning_rate": 9.913381348753213e-05, "loss": 0.0508, "step": 17020 }, { "epoch": 0.3403141361256545, "grad_norm": 0.15313690900802612, "learning_rate": 9.913189719943204e-05, "loss": 0.047, "step": 17030 }, { "epoch": 0.34051396826665603, "grad_norm": 0.14397859573364258, "learning_rate": 9.912997881250757e-05, "loss": 0.0451, "step": 17040 }, { "epoch": 0.3407138004076576, "grad_norm": 0.18657676875591278, "learning_rate": 9.912805832684066e-05, "loss": 0.0463, "step": 17050 }, { "epoch": 0.34091363254865914, "grad_norm": 0.14195559918880463, "learning_rate": 9.912613574251335e-05, "loss": 0.0418, "step": 17060 }, { "epoch": 0.3411134646896607, "grad_norm": 0.13905276358127594, "learning_rate": 9.912421105960776e-05, "loss": 0.0396, "step": 17070 }, { "epoch": 0.34131329683066225, "grad_norm": 0.19632762670516968, "learning_rate": 9.912228427820614e-05, "loss": 0.0432, "step": 17080 }, { "epoch": 0.3415131289716638, "grad_norm": 0.14380978047847748, "learning_rate": 9.912035539839076e-05, "loss": 0.0544, "step": 17090 }, { "epoch": 0.34171296111266536, "grad_norm": 0.1915758103132248, "learning_rate": 9.911842442024403e-05, "loss": 0.0464, "step": 17100 }, { "epoch": 0.3419127932536669, "grad_norm": 0.1312311589717865, "learning_rate": 9.911649134384845e-05, "loss": 0.0456, "step": 17110 }, { "epoch": 0.34211262539466847, "grad_norm": 0.15112537145614624, "learning_rate": 9.911455616928658e-05, "loss": 0.0385, "step": 17120 }, { "epoch": 0.34231245753567, "grad_norm": 0.19272008538246155, "learning_rate": 9.911261889664111e-05, "loss": 0.0487, "step": 17130 }, { "epoch": 0.3425122896766716, "grad_norm": 0.16099053621292114, "learning_rate": 9.911067952599478e-05, "loss": 0.053, "step": 17140 }, { "epoch": 0.34271212181767313, "grad_norm": 0.16505025327205658, "learning_rate": 9.910873805743044e-05, "loss": 0.0398, "step": 17150 }, { "epoch": 0.3429119539586747, "grad_norm": 0.21053659915924072, "learning_rate": 9.910679449103103e-05, "loss": 0.0476, "step": 17160 }, { "epoch": 0.3431117860996763, "grad_norm": 0.15500608086585999, "learning_rate": 9.910484882687956e-05, "loss": 0.0379, "step": 17170 }, { "epoch": 0.34331161824067785, "grad_norm": 0.17450197041034698, "learning_rate": 9.910290106505916e-05, "loss": 0.0521, "step": 17180 }, { "epoch": 0.3435114503816794, "grad_norm": 0.13368411362171173, "learning_rate": 9.910095120565303e-05, "loss": 0.0538, "step": 17190 }, { "epoch": 0.34371128252268096, "grad_norm": 0.1685383915901184, "learning_rate": 9.909899924874446e-05, "loss": 0.0522, "step": 17200 }, { "epoch": 0.3439111146636825, "grad_norm": 0.17234337329864502, "learning_rate": 9.909704519441686e-05, "loss": 0.0463, "step": 17210 }, { "epoch": 0.34411094680468407, "grad_norm": 0.1812209188938141, "learning_rate": 9.909508904275367e-05, "loss": 0.048, "step": 17220 }, { "epoch": 0.3443107789456856, "grad_norm": 0.1509702354669571, "learning_rate": 9.909313079383847e-05, "loss": 0.0456, "step": 17230 }, { "epoch": 0.3445106110866872, "grad_norm": 0.17195068299770355, "learning_rate": 9.909117044775491e-05, "loss": 0.0507, "step": 17240 }, { "epoch": 0.34471044322768873, "grad_norm": 0.13492487370967865, "learning_rate": 9.908920800458673e-05, "loss": 0.045, "step": 17250 }, { "epoch": 0.3449102753686903, "grad_norm": 0.14398293197155, "learning_rate": 9.908724346441776e-05, "loss": 0.0468, "step": 17260 }, { "epoch": 0.34511010750969184, "grad_norm": 0.16872581839561462, "learning_rate": 9.908527682733194e-05, "loss": 0.0496, "step": 17270 }, { "epoch": 0.3453099396506934, "grad_norm": 0.1391899436712265, "learning_rate": 9.908330809341324e-05, "loss": 0.04, "step": 17280 }, { "epoch": 0.34550977179169495, "grad_norm": 0.15160048007965088, "learning_rate": 9.90813372627458e-05, "loss": 0.0506, "step": 17290 }, { "epoch": 0.3457096039326965, "grad_norm": 0.1593182533979416, "learning_rate": 9.90793643354138e-05, "loss": 0.0459, "step": 17300 }, { "epoch": 0.3459094360736981, "grad_norm": 0.12621831893920898, "learning_rate": 9.907738931150152e-05, "loss": 0.0425, "step": 17310 }, { "epoch": 0.3461092682146997, "grad_norm": 0.17912086844444275, "learning_rate": 9.907541219109333e-05, "loss": 0.0487, "step": 17320 }, { "epoch": 0.3463091003557012, "grad_norm": 0.1889280080795288, "learning_rate": 9.907343297427367e-05, "loss": 0.0432, "step": 17330 }, { "epoch": 0.3465089324967028, "grad_norm": 0.147251158952713, "learning_rate": 9.907145166112712e-05, "loss": 0.0447, "step": 17340 }, { "epoch": 0.34670876463770434, "grad_norm": 0.1645045429468155, "learning_rate": 9.90694682517383e-05, "loss": 0.0524, "step": 17350 }, { "epoch": 0.3469085967787059, "grad_norm": 0.1567268669605255, "learning_rate": 9.906748274619193e-05, "loss": 0.0527, "step": 17360 }, { "epoch": 0.34710842891970745, "grad_norm": 0.13288496434688568, "learning_rate": 9.906549514457285e-05, "loss": 0.0434, "step": 17370 }, { "epoch": 0.347308261060709, "grad_norm": 0.21377572417259216, "learning_rate": 9.906350544696595e-05, "loss": 0.0485, "step": 17380 }, { "epoch": 0.34750809320171056, "grad_norm": 0.13127461075782776, "learning_rate": 9.906151365345621e-05, "loss": 0.0477, "step": 17390 }, { "epoch": 0.3477079253427121, "grad_norm": 0.11708460003137589, "learning_rate": 9.905951976412877e-05, "loss": 0.0405, "step": 17400 }, { "epoch": 0.34790775748371366, "grad_norm": 0.1498032808303833, "learning_rate": 9.905752377906874e-05, "loss": 0.0419, "step": 17410 }, { "epoch": 0.3481075896247152, "grad_norm": 0.15013328194618225, "learning_rate": 9.905552569836144e-05, "loss": 0.0407, "step": 17420 }, { "epoch": 0.3483074217657168, "grad_norm": 0.13249510526657104, "learning_rate": 9.905352552209217e-05, "loss": 0.0515, "step": 17430 }, { "epoch": 0.3485072539067184, "grad_norm": 0.13342982530593872, "learning_rate": 9.905152325034644e-05, "loss": 0.0439, "step": 17440 }, { "epoch": 0.34870708604771994, "grad_norm": 0.14733624458312988, "learning_rate": 9.904951888320972e-05, "loss": 0.0506, "step": 17450 }, { "epoch": 0.3489069181887215, "grad_norm": 0.13355526328086853, "learning_rate": 9.904751242076766e-05, "loss": 0.044, "step": 17460 }, { "epoch": 0.34910675032972305, "grad_norm": 0.16963738203048706, "learning_rate": 9.9045503863106e-05, "loss": 0.0405, "step": 17470 }, { "epoch": 0.3493065824707246, "grad_norm": 0.1658412367105484, "learning_rate": 9.904349321031049e-05, "loss": 0.0502, "step": 17480 }, { "epoch": 0.34950641461172616, "grad_norm": 0.17237898707389832, "learning_rate": 9.904148046246706e-05, "loss": 0.0503, "step": 17490 }, { "epoch": 0.3497062467527277, "grad_norm": 0.13741670548915863, "learning_rate": 9.903946561966167e-05, "loss": 0.0446, "step": 17500 }, { "epoch": 0.34990607889372927, "grad_norm": 0.1533478945493698, "learning_rate": 9.90374486819804e-05, "loss": 0.0472, "step": 17510 }, { "epoch": 0.3501059110347308, "grad_norm": 0.1587952822446823, "learning_rate": 9.903542964950939e-05, "loss": 0.0572, "step": 17520 }, { "epoch": 0.3503057431757324, "grad_norm": 0.1581016331911087, "learning_rate": 9.903340852233491e-05, "loss": 0.0432, "step": 17530 }, { "epoch": 0.35050557531673393, "grad_norm": 0.14366336166858673, "learning_rate": 9.903138530054332e-05, "loss": 0.0471, "step": 17540 }, { "epoch": 0.3507054074577355, "grad_norm": 0.15960685908794403, "learning_rate": 9.9029359984221e-05, "loss": 0.0494, "step": 17550 }, { "epoch": 0.35090523959873704, "grad_norm": 0.21138402819633484, "learning_rate": 9.90273325734545e-05, "loss": 0.0476, "step": 17560 }, { "epoch": 0.3511050717397386, "grad_norm": 0.1996658444404602, "learning_rate": 9.902530306833042e-05, "loss": 0.0677, "step": 17570 }, { "epoch": 0.3513049038807402, "grad_norm": 0.177456334233284, "learning_rate": 9.902327146893546e-05, "loss": 0.0409, "step": 17580 }, { "epoch": 0.35150473602174176, "grad_norm": 0.1772899478673935, "learning_rate": 9.90212377753564e-05, "loss": 0.0462, "step": 17590 }, { "epoch": 0.3517045681627433, "grad_norm": 0.15901033580303192, "learning_rate": 9.90192019876801e-05, "loss": 0.0483, "step": 17600 }, { "epoch": 0.35190440030374487, "grad_norm": 0.19647973775863647, "learning_rate": 9.901716410599356e-05, "loss": 0.0551, "step": 17610 }, { "epoch": 0.3521042324447464, "grad_norm": 0.16216909885406494, "learning_rate": 9.90151241303838e-05, "loss": 0.0442, "step": 17620 }, { "epoch": 0.352304064585748, "grad_norm": 0.15412697196006775, "learning_rate": 9.9013082060938e-05, "loss": 0.0414, "step": 17630 }, { "epoch": 0.35250389672674953, "grad_norm": 0.19386152923107147, "learning_rate": 9.901103789774337e-05, "loss": 0.0581, "step": 17640 }, { "epoch": 0.3527037288677511, "grad_norm": 0.19087037444114685, "learning_rate": 9.900899164088723e-05, "loss": 0.0477, "step": 17650 }, { "epoch": 0.35290356100875264, "grad_norm": 0.1684279888868332, "learning_rate": 9.9006943290457e-05, "loss": 0.0489, "step": 17660 }, { "epoch": 0.3531033931497542, "grad_norm": 0.11890789866447449, "learning_rate": 9.90048928465402e-05, "loss": 0.0407, "step": 17670 }, { "epoch": 0.35330322529075575, "grad_norm": 0.16974177956581116, "learning_rate": 9.900284030922437e-05, "loss": 0.0453, "step": 17680 }, { "epoch": 0.3535030574317573, "grad_norm": 0.15445969998836517, "learning_rate": 9.900078567859724e-05, "loss": 0.0455, "step": 17690 }, { "epoch": 0.35370288957275886, "grad_norm": 0.15719640254974365, "learning_rate": 9.899872895474656e-05, "loss": 0.0561, "step": 17700 }, { "epoch": 0.3539027217137604, "grad_norm": 0.1791938990354538, "learning_rate": 9.89966701377602e-05, "loss": 0.0516, "step": 17710 }, { "epoch": 0.354102553854762, "grad_norm": 0.17128923535346985, "learning_rate": 9.89946092277261e-05, "loss": 0.0502, "step": 17720 }, { "epoch": 0.3543023859957636, "grad_norm": 0.17147809267044067, "learning_rate": 9.899254622473229e-05, "loss": 0.0468, "step": 17730 }, { "epoch": 0.35450221813676513, "grad_norm": 0.17536379396915436, "learning_rate": 9.899048112886692e-05, "loss": 0.0368, "step": 17740 }, { "epoch": 0.3547020502777667, "grad_norm": 0.1353154480457306, "learning_rate": 9.898841394021818e-05, "loss": 0.0448, "step": 17750 }, { "epoch": 0.35490188241876824, "grad_norm": 0.16972415149211884, "learning_rate": 9.89863446588744e-05, "loss": 0.05, "step": 17760 }, { "epoch": 0.3551017145597698, "grad_norm": 0.1326778680086136, "learning_rate": 9.898427328492397e-05, "loss": 0.0444, "step": 17770 }, { "epoch": 0.35530154670077135, "grad_norm": 0.1675141304731369, "learning_rate": 9.898219981845537e-05, "loss": 0.0421, "step": 17780 }, { "epoch": 0.3555013788417729, "grad_norm": 0.1579221934080124, "learning_rate": 9.898012425955718e-05, "loss": 0.0411, "step": 17790 }, { "epoch": 0.35570121098277446, "grad_norm": 0.1636917144060135, "learning_rate": 9.897804660831805e-05, "loss": 0.0565, "step": 17800 }, { "epoch": 0.355901043123776, "grad_norm": 0.13882063329219818, "learning_rate": 9.897596686482675e-05, "loss": 0.042, "step": 17810 }, { "epoch": 0.35610087526477757, "grad_norm": 0.14901015162467957, "learning_rate": 9.897388502917212e-05, "loss": 0.0449, "step": 17820 }, { "epoch": 0.3563007074057791, "grad_norm": 0.1581115573644638, "learning_rate": 9.897180110144309e-05, "loss": 0.045, "step": 17830 }, { "epoch": 0.3565005395467807, "grad_norm": 0.13216592371463776, "learning_rate": 9.896971508172867e-05, "loss": 0.0461, "step": 17840 }, { "epoch": 0.35670037168778224, "grad_norm": 0.13592565059661865, "learning_rate": 9.8967626970118e-05, "loss": 0.0446, "step": 17850 }, { "epoch": 0.35690020382878385, "grad_norm": 0.1325157731771469, "learning_rate": 9.896553676670025e-05, "loss": 0.0385, "step": 17860 }, { "epoch": 0.3571000359697854, "grad_norm": 0.12266236543655396, "learning_rate": 9.896344447156473e-05, "loss": 0.0437, "step": 17870 }, { "epoch": 0.35729986811078696, "grad_norm": 0.15560804307460785, "learning_rate": 9.89613500848008e-05, "loss": 0.0449, "step": 17880 }, { "epoch": 0.3574997002517885, "grad_norm": 0.13008442521095276, "learning_rate": 9.895925360649794e-05, "loss": 0.0495, "step": 17890 }, { "epoch": 0.35769953239279006, "grad_norm": 0.12419208139181137, "learning_rate": 9.895715503674572e-05, "loss": 0.0447, "step": 17900 }, { "epoch": 0.3578993645337916, "grad_norm": 0.18485420942306519, "learning_rate": 9.895505437563377e-05, "loss": 0.0471, "step": 17910 }, { "epoch": 0.3580991966747932, "grad_norm": 0.19429858028888702, "learning_rate": 9.895295162325182e-05, "loss": 0.0509, "step": 17920 }, { "epoch": 0.35829902881579473, "grad_norm": 0.13059963285923004, "learning_rate": 9.895084677968971e-05, "loss": 0.0399, "step": 17930 }, { "epoch": 0.3584988609567963, "grad_norm": 0.14388535916805267, "learning_rate": 9.894873984503737e-05, "loss": 0.0426, "step": 17940 }, { "epoch": 0.35869869309779784, "grad_norm": 0.1435924917459488, "learning_rate": 9.894663081938475e-05, "loss": 0.0398, "step": 17950 }, { "epoch": 0.3588985252387994, "grad_norm": 0.13194479048252106, "learning_rate": 9.894451970282202e-05, "loss": 0.0512, "step": 17960 }, { "epoch": 0.35909835737980095, "grad_norm": 0.1503617763519287, "learning_rate": 9.89424064954393e-05, "loss": 0.0493, "step": 17970 }, { "epoch": 0.3592981895208025, "grad_norm": 0.20939628779888153, "learning_rate": 9.894029119732687e-05, "loss": 0.0466, "step": 17980 }, { "epoch": 0.35949802166180406, "grad_norm": 0.18472634255886078, "learning_rate": 9.893817380857512e-05, "loss": 0.0571, "step": 17990 }, { "epoch": 0.35969785380280567, "grad_norm": 0.18562740087509155, "learning_rate": 9.89360543292745e-05, "loss": 0.0478, "step": 18000 }, { "epoch": 0.3598976859438072, "grad_norm": 0.19032840430736542, "learning_rate": 9.893393275951552e-05, "loss": 0.0445, "step": 18010 }, { "epoch": 0.3600975180848088, "grad_norm": 0.18414966762065887, "learning_rate": 9.893180909938884e-05, "loss": 0.0431, "step": 18020 }, { "epoch": 0.36029735022581033, "grad_norm": 0.15722106397151947, "learning_rate": 9.892968334898516e-05, "loss": 0.0395, "step": 18030 }, { "epoch": 0.3604971823668119, "grad_norm": 0.1569712907075882, "learning_rate": 9.892755550839529e-05, "loss": 0.0418, "step": 18040 }, { "epoch": 0.36069701450781344, "grad_norm": 0.1419156789779663, "learning_rate": 9.892542557771014e-05, "loss": 0.0506, "step": 18050 }, { "epoch": 0.360896846648815, "grad_norm": 0.1718239188194275, "learning_rate": 9.892329355702068e-05, "loss": 0.0359, "step": 18060 }, { "epoch": 0.36109667878981655, "grad_norm": 0.1251058727502823, "learning_rate": 9.8921159446418e-05, "loss": 0.0406, "step": 18070 }, { "epoch": 0.3612965109308181, "grad_norm": 0.1556204855442047, "learning_rate": 9.891902324599324e-05, "loss": 0.0436, "step": 18080 }, { "epoch": 0.36149634307181966, "grad_norm": 0.14324714243412018, "learning_rate": 9.891688495583771e-05, "loss": 0.0405, "step": 18090 }, { "epoch": 0.3616961752128212, "grad_norm": 0.14841116964817047, "learning_rate": 9.891474457604269e-05, "loss": 0.0474, "step": 18100 }, { "epoch": 0.36189600735382277, "grad_norm": 0.14449721574783325, "learning_rate": 9.891260210669965e-05, "loss": 0.0489, "step": 18110 }, { "epoch": 0.3620958394948243, "grad_norm": 0.11768384277820587, "learning_rate": 9.89104575479001e-05, "loss": 0.0376, "step": 18120 }, { "epoch": 0.36229567163582593, "grad_norm": 0.1488465964794159, "learning_rate": 9.890831089973565e-05, "loss": 0.0415, "step": 18130 }, { "epoch": 0.3624955037768275, "grad_norm": 0.1727180778980255, "learning_rate": 9.890616216229803e-05, "loss": 0.0482, "step": 18140 }, { "epoch": 0.36269533591782904, "grad_norm": 0.12745168805122375, "learning_rate": 9.890401133567898e-05, "loss": 0.0449, "step": 18150 }, { "epoch": 0.3628951680588306, "grad_norm": 0.13572397828102112, "learning_rate": 9.890185841997042e-05, "loss": 0.0407, "step": 18160 }, { "epoch": 0.36309500019983215, "grad_norm": 0.14362101256847382, "learning_rate": 9.889970341526428e-05, "loss": 0.042, "step": 18170 }, { "epoch": 0.3632948323408337, "grad_norm": 0.16051289439201355, "learning_rate": 9.889754632165267e-05, "loss": 0.0496, "step": 18180 }, { "epoch": 0.36349466448183526, "grad_norm": 0.13536123931407928, "learning_rate": 9.88953871392277e-05, "loss": 0.0433, "step": 18190 }, { "epoch": 0.3636944966228368, "grad_norm": 0.20873522758483887, "learning_rate": 9.889322586808162e-05, "loss": 0.0476, "step": 18200 }, { "epoch": 0.36389432876383837, "grad_norm": 0.16081726551055908, "learning_rate": 9.889106250830674e-05, "loss": 0.0434, "step": 18210 }, { "epoch": 0.3640941609048399, "grad_norm": 0.1586841344833374, "learning_rate": 9.888889705999549e-05, "loss": 0.0543, "step": 18220 }, { "epoch": 0.3642939930458415, "grad_norm": 0.20047679543495178, "learning_rate": 9.888672952324038e-05, "loss": 0.0536, "step": 18230 }, { "epoch": 0.36449382518684303, "grad_norm": 0.15945309400558472, "learning_rate": 9.888455989813398e-05, "loss": 0.0436, "step": 18240 }, { "epoch": 0.3646936573278446, "grad_norm": 0.1264604777097702, "learning_rate": 9.888238818476898e-05, "loss": 0.0353, "step": 18250 }, { "epoch": 0.36489348946884614, "grad_norm": 0.14300478994846344, "learning_rate": 9.888021438323817e-05, "loss": 0.0385, "step": 18260 }, { "epoch": 0.36509332160984775, "grad_norm": 0.15977750718593597, "learning_rate": 9.88780384936344e-05, "loss": 0.0462, "step": 18270 }, { "epoch": 0.3652931537508493, "grad_norm": 0.16950641572475433, "learning_rate": 9.88758605160506e-05, "loss": 0.0473, "step": 18280 }, { "epoch": 0.36549298589185086, "grad_norm": 0.1362975388765335, "learning_rate": 9.887368045057983e-05, "loss": 0.0476, "step": 18290 }, { "epoch": 0.3656928180328524, "grad_norm": 0.1986132562160492, "learning_rate": 9.887149829731523e-05, "loss": 0.0541, "step": 18300 }, { "epoch": 0.365892650173854, "grad_norm": 0.15783879160881042, "learning_rate": 9.886931405634998e-05, "loss": 0.0458, "step": 18310 }, { "epoch": 0.3660924823148555, "grad_norm": 0.19384896755218506, "learning_rate": 9.886712772777742e-05, "loss": 0.0498, "step": 18320 }, { "epoch": 0.3662923144558571, "grad_norm": 0.16174539923667908, "learning_rate": 9.886493931169094e-05, "loss": 0.0421, "step": 18330 }, { "epoch": 0.36649214659685864, "grad_norm": 0.17831642925739288, "learning_rate": 9.886274880818403e-05, "loss": 0.0563, "step": 18340 }, { "epoch": 0.3666919787378602, "grad_norm": 0.12081193923950195, "learning_rate": 9.886055621735023e-05, "loss": 0.0498, "step": 18350 }, { "epoch": 0.36689181087886175, "grad_norm": 0.1668892204761505, "learning_rate": 9.885836153928325e-05, "loss": 0.0478, "step": 18360 }, { "epoch": 0.3670916430198633, "grad_norm": 0.14862827956676483, "learning_rate": 9.885616477407679e-05, "loss": 0.0423, "step": 18370 }, { "epoch": 0.36729147516086486, "grad_norm": 0.16205371916294098, "learning_rate": 9.885396592182477e-05, "loss": 0.0431, "step": 18380 }, { "epoch": 0.3674913073018664, "grad_norm": 0.13976401090621948, "learning_rate": 9.885176498262104e-05, "loss": 0.0518, "step": 18390 }, { "epoch": 0.36769113944286796, "grad_norm": 0.15774036943912506, "learning_rate": 9.884956195655966e-05, "loss": 0.0423, "step": 18400 }, { "epoch": 0.3678909715838696, "grad_norm": 0.16117067635059357, "learning_rate": 9.884735684373473e-05, "loss": 0.0475, "step": 18410 }, { "epoch": 0.36809080372487113, "grad_norm": 0.14457520842552185, "learning_rate": 9.884514964424046e-05, "loss": 0.0446, "step": 18420 }, { "epoch": 0.3682906358658727, "grad_norm": 0.12450730055570602, "learning_rate": 9.884294035817112e-05, "loss": 0.0502, "step": 18430 }, { "epoch": 0.36849046800687424, "grad_norm": 0.12593264877796173, "learning_rate": 9.884072898562109e-05, "loss": 0.0473, "step": 18440 }, { "epoch": 0.3686903001478758, "grad_norm": 0.12088844180107117, "learning_rate": 9.883851552668486e-05, "loss": 0.046, "step": 18450 }, { "epoch": 0.36889013228887735, "grad_norm": 0.1720925271511078, "learning_rate": 9.883629998145695e-05, "loss": 0.0469, "step": 18460 }, { "epoch": 0.3690899644298789, "grad_norm": 0.14555703103542328, "learning_rate": 9.883408235003202e-05, "loss": 0.051, "step": 18470 }, { "epoch": 0.36928979657088046, "grad_norm": 0.12528884410858154, "learning_rate": 9.883186263250482e-05, "loss": 0.0387, "step": 18480 }, { "epoch": 0.369489628711882, "grad_norm": 0.1343759000301361, "learning_rate": 9.882964082897013e-05, "loss": 0.0508, "step": 18490 }, { "epoch": 0.36968946085288357, "grad_norm": 0.1415061503648758, "learning_rate": 9.882741693952291e-05, "loss": 0.0436, "step": 18500 }, { "epoch": 0.3698892929938851, "grad_norm": 0.14968141913414001, "learning_rate": 9.882519096425814e-05, "loss": 0.0484, "step": 18510 }, { "epoch": 0.3700891251348867, "grad_norm": 0.11596011370420456, "learning_rate": 9.882296290327088e-05, "loss": 0.0412, "step": 18520 }, { "epoch": 0.37028895727588823, "grad_norm": 0.14935435354709625, "learning_rate": 9.882073275665636e-05, "loss": 0.0515, "step": 18530 }, { "epoch": 0.3704887894168898, "grad_norm": 0.16596807539463043, "learning_rate": 9.881850052450981e-05, "loss": 0.0447, "step": 18540 }, { "epoch": 0.3706886215578914, "grad_norm": 0.14093057811260223, "learning_rate": 9.881626620692659e-05, "loss": 0.0398, "step": 18550 }, { "epoch": 0.37088845369889295, "grad_norm": 0.15055438876152039, "learning_rate": 9.881402980400217e-05, "loss": 0.0421, "step": 18560 }, { "epoch": 0.3710882858398945, "grad_norm": 0.131365105509758, "learning_rate": 9.881179131583207e-05, "loss": 0.043, "step": 18570 }, { "epoch": 0.37128811798089606, "grad_norm": 0.18391841650009155, "learning_rate": 9.880955074251191e-05, "loss": 0.0381, "step": 18580 }, { "epoch": 0.3714879501218976, "grad_norm": 0.16265486180782318, "learning_rate": 9.880730808413742e-05, "loss": 0.0483, "step": 18590 }, { "epoch": 0.37168778226289917, "grad_norm": 0.1971440613269806, "learning_rate": 9.880506334080438e-05, "loss": 0.0454, "step": 18600 }, { "epoch": 0.3718876144039007, "grad_norm": 0.15393628180027008, "learning_rate": 9.88028165126087e-05, "loss": 0.0417, "step": 18610 }, { "epoch": 0.3720874465449023, "grad_norm": 0.12183696031570435, "learning_rate": 9.880056759964634e-05, "loss": 0.0466, "step": 18620 }, { "epoch": 0.37228727868590383, "grad_norm": 0.1823771595954895, "learning_rate": 9.87983166020134e-05, "loss": 0.0429, "step": 18630 }, { "epoch": 0.3724871108269054, "grad_norm": 0.14285430312156677, "learning_rate": 9.8796063519806e-05, "loss": 0.05, "step": 18640 }, { "epoch": 0.37268694296790694, "grad_norm": 0.12280312180519104, "learning_rate": 9.879380835312042e-05, "loss": 0.0425, "step": 18650 }, { "epoch": 0.3728867751089085, "grad_norm": 0.14697027206420898, "learning_rate": 9.879155110205299e-05, "loss": 0.0594, "step": 18660 }, { "epoch": 0.37308660724991005, "grad_norm": 0.15276610851287842, "learning_rate": 9.878929176670011e-05, "loss": 0.0509, "step": 18670 }, { "epoch": 0.37328643939091166, "grad_norm": 0.1949511170387268, "learning_rate": 9.878703034715833e-05, "loss": 0.0443, "step": 18680 }, { "epoch": 0.3734862715319132, "grad_norm": 0.1360844075679779, "learning_rate": 9.878476684352424e-05, "loss": 0.0425, "step": 18690 }, { "epoch": 0.37368610367291477, "grad_norm": 0.14476512372493744, "learning_rate": 9.878250125589452e-05, "loss": 0.0391, "step": 18700 }, { "epoch": 0.3738859358139163, "grad_norm": 0.15296852588653564, "learning_rate": 9.878023358436596e-05, "loss": 0.0479, "step": 18710 }, { "epoch": 0.3740857679549179, "grad_norm": 0.14366470277309418, "learning_rate": 9.877796382903544e-05, "loss": 0.0459, "step": 18720 }, { "epoch": 0.37428560009591944, "grad_norm": 0.14209936559200287, "learning_rate": 9.877569198999992e-05, "loss": 0.0498, "step": 18730 }, { "epoch": 0.374485432236921, "grad_norm": 0.15189290046691895, "learning_rate": 9.877341806735643e-05, "loss": 0.0482, "step": 18740 }, { "epoch": 0.37468526437792254, "grad_norm": 0.16466790437698364, "learning_rate": 9.877114206120212e-05, "loss": 0.044, "step": 18750 }, { "epoch": 0.3748850965189241, "grad_norm": 0.14442965388298035, "learning_rate": 9.876886397163423e-05, "loss": 0.0533, "step": 18760 }, { "epoch": 0.37508492865992565, "grad_norm": 0.16528882086277008, "learning_rate": 9.876658379875006e-05, "loss": 0.0465, "step": 18770 }, { "epoch": 0.3752847608009272, "grad_norm": 0.15622173249721527, "learning_rate": 9.876430154264702e-05, "loss": 0.0483, "step": 18780 }, { "epoch": 0.37548459294192876, "grad_norm": 0.14181287586688995, "learning_rate": 9.876201720342258e-05, "loss": 0.0386, "step": 18790 }, { "epoch": 0.3756844250829303, "grad_norm": 0.13109305500984192, "learning_rate": 9.875973078117436e-05, "loss": 0.0439, "step": 18800 }, { "epoch": 0.3758842572239319, "grad_norm": 0.12001454830169678, "learning_rate": 9.8757442276e-05, "loss": 0.0413, "step": 18810 }, { "epoch": 0.3760840893649335, "grad_norm": 0.1313575655221939, "learning_rate": 9.87551516879973e-05, "loss": 0.0449, "step": 18820 }, { "epoch": 0.37628392150593504, "grad_norm": 0.16821913421154022, "learning_rate": 9.875285901726408e-05, "loss": 0.0519, "step": 18830 }, { "epoch": 0.3764837536469366, "grad_norm": 0.1291782110929489, "learning_rate": 9.875056426389828e-05, "loss": 0.0392, "step": 18840 }, { "epoch": 0.37668358578793815, "grad_norm": 0.1476517617702484, "learning_rate": 9.874826742799793e-05, "loss": 0.0453, "step": 18850 }, { "epoch": 0.3768834179289397, "grad_norm": 0.17535454034805298, "learning_rate": 9.874596850966116e-05, "loss": 0.044, "step": 18860 }, { "epoch": 0.37708325006994126, "grad_norm": 0.17056432366371155, "learning_rate": 9.874366750898615e-05, "loss": 0.0372, "step": 18870 }, { "epoch": 0.3772830822109428, "grad_norm": 0.14573262631893158, "learning_rate": 9.87413644260712e-05, "loss": 0.0615, "step": 18880 }, { "epoch": 0.37748291435194437, "grad_norm": 0.13404135406017303, "learning_rate": 9.873905926101472e-05, "loss": 0.0477, "step": 18890 }, { "epoch": 0.3776827464929459, "grad_norm": 0.13413460552692413, "learning_rate": 9.873675201391517e-05, "loss": 0.0389, "step": 18900 }, { "epoch": 0.3778825786339475, "grad_norm": 0.1477365791797638, "learning_rate": 9.87344426848711e-05, "loss": 0.0477, "step": 18910 }, { "epoch": 0.37808241077494903, "grad_norm": 0.1551104187965393, "learning_rate": 9.873213127398115e-05, "loss": 0.048, "step": 18920 }, { "epoch": 0.3782822429159506, "grad_norm": 0.1553383320569992, "learning_rate": 9.872981778134409e-05, "loss": 0.0445, "step": 18930 }, { "epoch": 0.37848207505695214, "grad_norm": 0.11509757488965988, "learning_rate": 9.872750220705874e-05, "loss": 0.0467, "step": 18940 }, { "epoch": 0.3786819071979537, "grad_norm": 0.14194256067276, "learning_rate": 9.872518455122401e-05, "loss": 0.0427, "step": 18950 }, { "epoch": 0.3788817393389553, "grad_norm": 0.16332995891571045, "learning_rate": 9.87228648139389e-05, "loss": 0.0454, "step": 18960 }, { "epoch": 0.37908157147995686, "grad_norm": 0.1382971853017807, "learning_rate": 9.87205429953025e-05, "loss": 0.0477, "step": 18970 }, { "epoch": 0.3792814036209584, "grad_norm": 0.1335146278142929, "learning_rate": 9.871821909541404e-05, "loss": 0.0529, "step": 18980 }, { "epoch": 0.37948123576195997, "grad_norm": 0.1184939295053482, "learning_rate": 9.871589311437272e-05, "loss": 0.0411, "step": 18990 }, { "epoch": 0.3796810679029615, "grad_norm": 0.17756465077400208, "learning_rate": 9.871356505227797e-05, "loss": 0.0528, "step": 19000 }, { "epoch": 0.3798809000439631, "grad_norm": 0.1359984129667282, "learning_rate": 9.87112349092292e-05, "loss": 0.0448, "step": 19010 }, { "epoch": 0.38008073218496463, "grad_norm": 0.16830909252166748, "learning_rate": 9.870890268532596e-05, "loss": 0.0428, "step": 19020 }, { "epoch": 0.3802805643259662, "grad_norm": 0.14912381768226624, "learning_rate": 9.870656838066788e-05, "loss": 0.0526, "step": 19030 }, { "epoch": 0.38048039646696774, "grad_norm": 0.18155434727668762, "learning_rate": 9.870423199535467e-05, "loss": 0.0462, "step": 19040 }, { "epoch": 0.3806802286079693, "grad_norm": 0.1636820137500763, "learning_rate": 9.870189352948614e-05, "loss": 0.0461, "step": 19050 }, { "epoch": 0.38088006074897085, "grad_norm": 0.13440380990505219, "learning_rate": 9.869955298316219e-05, "loss": 0.0497, "step": 19060 }, { "epoch": 0.3810798928899724, "grad_norm": 0.13060882687568665, "learning_rate": 9.869721035648281e-05, "loss": 0.0547, "step": 19070 }, { "epoch": 0.38127972503097396, "grad_norm": 0.1535608470439911, "learning_rate": 9.869486564954804e-05, "loss": 0.0427, "step": 19080 }, { "epoch": 0.3814795571719755, "grad_norm": 0.17098930478096008, "learning_rate": 9.86925188624581e-05, "loss": 0.0503, "step": 19090 }, { "epoch": 0.3816793893129771, "grad_norm": 0.13536031544208527, "learning_rate": 9.869016999531317e-05, "loss": 0.0493, "step": 19100 }, { "epoch": 0.3818792214539787, "grad_norm": 0.16353586316108704, "learning_rate": 9.868781904821365e-05, "loss": 0.0446, "step": 19110 }, { "epoch": 0.38207905359498023, "grad_norm": 0.10344463586807251, "learning_rate": 9.868546602125992e-05, "loss": 0.0544, "step": 19120 }, { "epoch": 0.3822788857359818, "grad_norm": 0.14913171529769897, "learning_rate": 9.868311091455253e-05, "loss": 0.0408, "step": 19130 }, { "epoch": 0.38247871787698334, "grad_norm": 0.1504240483045578, "learning_rate": 9.868075372819208e-05, "loss": 0.0535, "step": 19140 }, { "epoch": 0.3826785500179849, "grad_norm": 0.12622252106666565, "learning_rate": 9.867839446227927e-05, "loss": 0.0459, "step": 19150 }, { "epoch": 0.38287838215898645, "grad_norm": 0.14335089921951294, "learning_rate": 9.867603311691485e-05, "loss": 0.0493, "step": 19160 }, { "epoch": 0.383078214299988, "grad_norm": 0.14928656816482544, "learning_rate": 9.867366969219973e-05, "loss": 0.0465, "step": 19170 }, { "epoch": 0.38327804644098956, "grad_norm": 0.11203356087207794, "learning_rate": 9.867130418823485e-05, "loss": 0.0333, "step": 19180 }, { "epoch": 0.3834778785819911, "grad_norm": 0.1471448838710785, "learning_rate": 9.866893660512127e-05, "loss": 0.0435, "step": 19190 }, { "epoch": 0.38367771072299267, "grad_norm": 0.17533011734485626, "learning_rate": 9.866656694296013e-05, "loss": 0.0496, "step": 19200 }, { "epoch": 0.3838775428639942, "grad_norm": 0.17385753989219666, "learning_rate": 9.866419520185265e-05, "loss": 0.0521, "step": 19210 }, { "epoch": 0.3840773750049958, "grad_norm": 0.18527162075042725, "learning_rate": 9.866182138190014e-05, "loss": 0.0524, "step": 19220 }, { "epoch": 0.38427720714599733, "grad_norm": 0.1751960813999176, "learning_rate": 9.865944548320404e-05, "loss": 0.0505, "step": 19230 }, { "epoch": 0.38447703928699895, "grad_norm": 0.1400946080684662, "learning_rate": 9.865706750586581e-05, "loss": 0.053, "step": 19240 }, { "epoch": 0.3846768714280005, "grad_norm": 0.14067724347114563, "learning_rate": 9.865468744998703e-05, "loss": 0.0527, "step": 19250 }, { "epoch": 0.38487670356900205, "grad_norm": 0.1563810110092163, "learning_rate": 9.86523053156694e-05, "loss": 0.0454, "step": 19260 }, { "epoch": 0.3850765357100036, "grad_norm": 0.1487371027469635, "learning_rate": 9.864992110301465e-05, "loss": 0.0434, "step": 19270 }, { "epoch": 0.38527636785100516, "grad_norm": 0.16113992035388947, "learning_rate": 9.864753481212465e-05, "loss": 0.0389, "step": 19280 }, { "epoch": 0.3854761999920067, "grad_norm": 0.15331897139549255, "learning_rate": 9.864514644310133e-05, "loss": 0.0501, "step": 19290 }, { "epoch": 0.3856760321330083, "grad_norm": 0.15534202754497528, "learning_rate": 9.864275599604671e-05, "loss": 0.0431, "step": 19300 }, { "epoch": 0.3858758642740098, "grad_norm": 0.17724592983722687, "learning_rate": 9.864036347106292e-05, "loss": 0.0417, "step": 19310 }, { "epoch": 0.3860756964150114, "grad_norm": 0.1402415931224823, "learning_rate": 9.863796886825216e-05, "loss": 0.0497, "step": 19320 }, { "epoch": 0.38627552855601294, "grad_norm": 0.1165580078959465, "learning_rate": 9.86355721877167e-05, "loss": 0.0451, "step": 19330 }, { "epoch": 0.3864753606970145, "grad_norm": 0.16446250677108765, "learning_rate": 9.863317342955898e-05, "loss": 0.0486, "step": 19340 }, { "epoch": 0.38667519283801605, "grad_norm": 0.19190818071365356, "learning_rate": 9.86307725938814e-05, "loss": 0.048, "step": 19350 }, { "epoch": 0.3868750249790176, "grad_norm": 0.1730780452489853, "learning_rate": 9.862836968078658e-05, "loss": 0.0492, "step": 19360 }, { "epoch": 0.3870748571200192, "grad_norm": 0.17509669065475464, "learning_rate": 9.862596469037713e-05, "loss": 0.0413, "step": 19370 }, { "epoch": 0.38727468926102077, "grad_norm": 0.14700311422348022, "learning_rate": 9.862355762275578e-05, "loss": 0.0436, "step": 19380 }, { "epoch": 0.3874745214020223, "grad_norm": 0.12863405048847198, "learning_rate": 9.86211484780254e-05, "loss": 0.0476, "step": 19390 }, { "epoch": 0.3876743535430239, "grad_norm": 0.130859836935997, "learning_rate": 9.861873725628887e-05, "loss": 0.041, "step": 19400 }, { "epoch": 0.38787418568402543, "grad_norm": 0.13969774544239044, "learning_rate": 9.861632395764919e-05, "loss": 0.0492, "step": 19410 }, { "epoch": 0.388074017825027, "grad_norm": 0.15911568701267242, "learning_rate": 9.861390858220947e-05, "loss": 0.0484, "step": 19420 }, { "epoch": 0.38827384996602854, "grad_norm": 0.18383529782295227, "learning_rate": 9.861149113007286e-05, "loss": 0.0443, "step": 19430 }, { "epoch": 0.3884736821070301, "grad_norm": 0.11782217770814896, "learning_rate": 9.860907160134267e-05, "loss": 0.0446, "step": 19440 }, { "epoch": 0.38867351424803165, "grad_norm": 0.12683163583278656, "learning_rate": 9.860664999612224e-05, "loss": 0.0436, "step": 19450 }, { "epoch": 0.3888733463890332, "grad_norm": 0.15135590732097626, "learning_rate": 9.860422631451503e-05, "loss": 0.0538, "step": 19460 }, { "epoch": 0.38907317853003476, "grad_norm": 0.12400764226913452, "learning_rate": 9.860180055662453e-05, "loss": 0.0478, "step": 19470 }, { "epoch": 0.3892730106710363, "grad_norm": 0.14411552250385284, "learning_rate": 9.859937272255442e-05, "loss": 0.0401, "step": 19480 }, { "epoch": 0.38947284281203787, "grad_norm": 0.19791574776172638, "learning_rate": 9.859694281240836e-05, "loss": 0.0484, "step": 19490 }, { "epoch": 0.3896726749530394, "grad_norm": 0.1412687599658966, "learning_rate": 9.85945108262902e-05, "loss": 0.0408, "step": 19500 }, { "epoch": 0.38987250709404103, "grad_norm": 0.15631873905658722, "learning_rate": 9.85920767643038e-05, "loss": 0.0427, "step": 19510 }, { "epoch": 0.3900723392350426, "grad_norm": 0.1602047085762024, "learning_rate": 9.858964062655316e-05, "loss": 0.0469, "step": 19520 }, { "epoch": 0.39027217137604414, "grad_norm": 0.122756227850914, "learning_rate": 9.858720241314231e-05, "loss": 0.0465, "step": 19530 }, { "epoch": 0.3904720035170457, "grad_norm": 0.21956129372119904, "learning_rate": 9.858476212417544e-05, "loss": 0.0429, "step": 19540 }, { "epoch": 0.39067183565804725, "grad_norm": 0.29439881443977356, "learning_rate": 9.85823197597568e-05, "loss": 0.0505, "step": 19550 }, { "epoch": 0.3908716677990488, "grad_norm": 0.1712574064731598, "learning_rate": 9.85798753199907e-05, "loss": 0.0396, "step": 19560 }, { "epoch": 0.39107149994005036, "grad_norm": 0.14765682816505432, "learning_rate": 9.857742880498157e-05, "loss": 0.0401, "step": 19570 }, { "epoch": 0.3912713320810519, "grad_norm": 0.14145927131175995, "learning_rate": 9.857498021483391e-05, "loss": 0.0532, "step": 19580 }, { "epoch": 0.39147116422205347, "grad_norm": 0.11238993704319, "learning_rate": 9.857252954965234e-05, "loss": 0.0386, "step": 19590 }, { "epoch": 0.391670996363055, "grad_norm": 0.13209028542041779, "learning_rate": 9.857007680954154e-05, "loss": 0.0433, "step": 19600 }, { "epoch": 0.3918708285040566, "grad_norm": 0.15456309914588928, "learning_rate": 9.856762199460629e-05, "loss": 0.0469, "step": 19610 }, { "epoch": 0.39207066064505813, "grad_norm": 0.1698312759399414, "learning_rate": 9.856516510495144e-05, "loss": 0.0477, "step": 19620 }, { "epoch": 0.3922704927860597, "grad_norm": 0.1640891134738922, "learning_rate": 9.856270614068195e-05, "loss": 0.0379, "step": 19630 }, { "epoch": 0.39247032492706124, "grad_norm": 0.18344052135944366, "learning_rate": 9.856024510190288e-05, "loss": 0.0449, "step": 19640 }, { "epoch": 0.39267015706806285, "grad_norm": 0.17114028334617615, "learning_rate": 9.855778198871934e-05, "loss": 0.0584, "step": 19650 }, { "epoch": 0.3928699892090644, "grad_norm": 0.12404689937829971, "learning_rate": 9.855531680123655e-05, "loss": 0.0453, "step": 19660 }, { "epoch": 0.39306982135006596, "grad_norm": 0.15226854383945465, "learning_rate": 9.855284953955984e-05, "loss": 0.046, "step": 19670 }, { "epoch": 0.3932696534910675, "grad_norm": 0.15489977598190308, "learning_rate": 9.855038020379458e-05, "loss": 0.0534, "step": 19680 }, { "epoch": 0.39346948563206907, "grad_norm": 0.22517111897468567, "learning_rate": 9.854790879404627e-05, "loss": 0.0477, "step": 19690 }, { "epoch": 0.3936693177730706, "grad_norm": 0.18080440163612366, "learning_rate": 9.854543531042048e-05, "loss": 0.0582, "step": 19700 }, { "epoch": 0.3938691499140722, "grad_norm": 0.1453384906053543, "learning_rate": 9.854295975302288e-05, "loss": 0.0473, "step": 19710 }, { "epoch": 0.39406898205507374, "grad_norm": 0.14640991389751434, "learning_rate": 9.85404821219592e-05, "loss": 0.0475, "step": 19720 }, { "epoch": 0.3942688141960753, "grad_norm": 0.13739469647407532, "learning_rate": 9.853800241733533e-05, "loss": 0.0419, "step": 19730 }, { "epoch": 0.39446864633707684, "grad_norm": 0.14527757465839386, "learning_rate": 9.853552063925713e-05, "loss": 0.0438, "step": 19740 }, { "epoch": 0.3946684784780784, "grad_norm": 0.12681013345718384, "learning_rate": 9.853303678783068e-05, "loss": 0.0491, "step": 19750 }, { "epoch": 0.39486831061907995, "grad_norm": 0.13669295608997345, "learning_rate": 9.853055086316203e-05, "loss": 0.044, "step": 19760 }, { "epoch": 0.3950681427600815, "grad_norm": 0.1325189769268036, "learning_rate": 9.852806286535741e-05, "loss": 0.0467, "step": 19770 }, { "epoch": 0.39526797490108306, "grad_norm": 0.16892023384571075, "learning_rate": 9.85255727945231e-05, "loss": 0.0569, "step": 19780 }, { "epoch": 0.3954678070420847, "grad_norm": 0.16738596558570862, "learning_rate": 9.852308065076545e-05, "loss": 0.0446, "step": 19790 }, { "epoch": 0.39566763918308623, "grad_norm": 0.1712912619113922, "learning_rate": 9.852058643419095e-05, "loss": 0.0509, "step": 19800 }, { "epoch": 0.3958674713240878, "grad_norm": 0.16720575094223022, "learning_rate": 9.851809014490613e-05, "loss": 0.0414, "step": 19810 }, { "epoch": 0.39606730346508934, "grad_norm": 0.12645982205867767, "learning_rate": 9.851559178301761e-05, "loss": 0.0463, "step": 19820 }, { "epoch": 0.3962671356060909, "grad_norm": 0.15845850110054016, "learning_rate": 9.851309134863216e-05, "loss": 0.0477, "step": 19830 }, { "epoch": 0.39646696774709245, "grad_norm": 0.14005368947982788, "learning_rate": 9.851058884185656e-05, "loss": 0.0426, "step": 19840 }, { "epoch": 0.396666799888094, "grad_norm": 0.16907283663749695, "learning_rate": 9.850808426279772e-05, "loss": 0.0363, "step": 19850 }, { "epoch": 0.39686663202909556, "grad_norm": 0.12848983705043793, "learning_rate": 9.850557761156263e-05, "loss": 0.0398, "step": 19860 }, { "epoch": 0.3970664641700971, "grad_norm": 0.1355930119752884, "learning_rate": 9.850306888825838e-05, "loss": 0.0585, "step": 19870 }, { "epoch": 0.39726629631109867, "grad_norm": 0.1400337815284729, "learning_rate": 9.850055809299212e-05, "loss": 0.0489, "step": 19880 }, { "epoch": 0.3974661284521002, "grad_norm": 0.14699532091617584, "learning_rate": 9.849804522587112e-05, "loss": 0.0501, "step": 19890 }, { "epoch": 0.3976659605931018, "grad_norm": 0.17260566353797913, "learning_rate": 9.849553028700273e-05, "loss": 0.0558, "step": 19900 }, { "epoch": 0.39786579273410333, "grad_norm": 0.1771761178970337, "learning_rate": 9.849301327649436e-05, "loss": 0.0472, "step": 19910 }, { "epoch": 0.3980656248751049, "grad_norm": 0.15060192346572876, "learning_rate": 9.849049419445356e-05, "loss": 0.0429, "step": 19920 }, { "epoch": 0.3982654570161065, "grad_norm": 0.17443281412124634, "learning_rate": 9.848797304098793e-05, "loss": 0.0417, "step": 19930 }, { "epoch": 0.39846528915710805, "grad_norm": 0.1581091284751892, "learning_rate": 9.848544981620516e-05, "loss": 0.051, "step": 19940 }, { "epoch": 0.3986651212981096, "grad_norm": 0.1641068011522293, "learning_rate": 9.848292452021305e-05, "loss": 0.0386, "step": 19950 }, { "epoch": 0.39886495343911116, "grad_norm": 0.333210825920105, "learning_rate": 9.848039715311947e-05, "loss": 0.0509, "step": 19960 }, { "epoch": 0.3990647855801127, "grad_norm": 0.1565418392419815, "learning_rate": 9.847786771503238e-05, "loss": 0.0466, "step": 19970 }, { "epoch": 0.39926461772111427, "grad_norm": 0.15456697344779968, "learning_rate": 9.847533620605983e-05, "loss": 0.0465, "step": 19980 }, { "epoch": 0.3994644498621158, "grad_norm": 0.15712684392929077, "learning_rate": 9.847280262630998e-05, "loss": 0.0502, "step": 19990 }, { "epoch": 0.3996642820031174, "grad_norm": 0.16654075682163239, "learning_rate": 9.847026697589106e-05, "loss": 0.0512, "step": 20000 }, { "epoch": 0.39986411414411893, "grad_norm": 0.15998709201812744, "learning_rate": 9.846772925491136e-05, "loss": 0.0466, "step": 20010 }, { "epoch": 0.4000639462851205, "grad_norm": 0.18335901200771332, "learning_rate": 9.84651894634793e-05, "loss": 0.0448, "step": 20020 }, { "epoch": 0.40026377842612204, "grad_norm": 0.19193552434444427, "learning_rate": 9.846264760170339e-05, "loss": 0.0477, "step": 20030 }, { "epoch": 0.4004636105671236, "grad_norm": 0.1634751260280609, "learning_rate": 9.84601036696922e-05, "loss": 0.0482, "step": 20040 }, { "epoch": 0.40066344270812515, "grad_norm": 0.15985073149204254, "learning_rate": 9.845755766755441e-05, "loss": 0.051, "step": 20050 }, { "epoch": 0.40086327484912676, "grad_norm": 0.17023852467536926, "learning_rate": 9.845500959539877e-05, "loss": 0.0415, "step": 20060 }, { "epoch": 0.4010631069901283, "grad_norm": 0.15123991668224335, "learning_rate": 9.845245945333414e-05, "loss": 0.0517, "step": 20070 }, { "epoch": 0.40126293913112987, "grad_norm": 0.17626869678497314, "learning_rate": 9.844990724146946e-05, "loss": 0.0572, "step": 20080 }, { "epoch": 0.4014627712721314, "grad_norm": 0.1279800683259964, "learning_rate": 9.844735295991373e-05, "loss": 0.0434, "step": 20090 }, { "epoch": 0.401662603413133, "grad_norm": 0.12360991537570953, "learning_rate": 9.844479660877609e-05, "loss": 0.0439, "step": 20100 }, { "epoch": 0.40186243555413453, "grad_norm": 0.19419358670711517, "learning_rate": 9.844223818816574e-05, "loss": 0.0535, "step": 20110 }, { "epoch": 0.4020622676951361, "grad_norm": 0.180254727602005, "learning_rate": 9.843967769819195e-05, "loss": 0.0456, "step": 20120 }, { "epoch": 0.40226209983613764, "grad_norm": 0.16149581968784332, "learning_rate": 9.843711513896412e-05, "loss": 0.0442, "step": 20130 }, { "epoch": 0.4024619319771392, "grad_norm": 0.14863449335098267, "learning_rate": 9.843455051059174e-05, "loss": 0.0438, "step": 20140 }, { "epoch": 0.40266176411814075, "grad_norm": 0.09934207051992416, "learning_rate": 9.84319838131843e-05, "loss": 0.0437, "step": 20150 }, { "epoch": 0.4028615962591423, "grad_norm": 0.1583290547132492, "learning_rate": 9.842941504685152e-05, "loss": 0.0439, "step": 20160 }, { "epoch": 0.40306142840014386, "grad_norm": 0.16006770730018616, "learning_rate": 9.842684421170308e-05, "loss": 0.0432, "step": 20170 }, { "epoch": 0.4032612605411454, "grad_norm": 0.16509319841861725, "learning_rate": 9.842427130784882e-05, "loss": 0.0405, "step": 20180 }, { "epoch": 0.40346109268214697, "grad_norm": 0.14289520680904388, "learning_rate": 9.842169633539864e-05, "loss": 0.0465, "step": 20190 }, { "epoch": 0.4036609248231486, "grad_norm": 0.14352644979953766, "learning_rate": 9.841911929446255e-05, "loss": 0.041, "step": 20200 }, { "epoch": 0.40386075696415014, "grad_norm": 0.14877162873744965, "learning_rate": 9.841654018515063e-05, "loss": 0.0448, "step": 20210 }, { "epoch": 0.4040605891051517, "grad_norm": 0.17060361802577972, "learning_rate": 9.841395900757307e-05, "loss": 0.0497, "step": 20220 }, { "epoch": 0.40426042124615325, "grad_norm": 0.16855765879154205, "learning_rate": 9.841137576184009e-05, "loss": 0.0415, "step": 20230 }, { "epoch": 0.4044602533871548, "grad_norm": 0.1477939933538437, "learning_rate": 9.84087904480621e-05, "loss": 0.0492, "step": 20240 }, { "epoch": 0.40466008552815635, "grad_norm": 0.1765286922454834, "learning_rate": 9.840620306634951e-05, "loss": 0.0392, "step": 20250 }, { "epoch": 0.4048599176691579, "grad_norm": 0.12752893567085266, "learning_rate": 9.840361361681285e-05, "loss": 0.045, "step": 20260 }, { "epoch": 0.40505974981015946, "grad_norm": 0.13883128762245178, "learning_rate": 9.840102209956272e-05, "loss": 0.0365, "step": 20270 }, { "epoch": 0.405259581951161, "grad_norm": 0.14695654809474945, "learning_rate": 9.839842851470986e-05, "loss": 0.0479, "step": 20280 }, { "epoch": 0.4054594140921626, "grad_norm": 0.18216535449028015, "learning_rate": 9.839583286236505e-05, "loss": 0.043, "step": 20290 }, { "epoch": 0.40565924623316413, "grad_norm": 0.14466838538646698, "learning_rate": 9.839323514263914e-05, "loss": 0.0451, "step": 20300 }, { "epoch": 0.4058590783741657, "grad_norm": 0.15963183343410492, "learning_rate": 9.839063535564315e-05, "loss": 0.0436, "step": 20310 }, { "epoch": 0.40605891051516724, "grad_norm": 0.1390843391418457, "learning_rate": 9.83880335014881e-05, "loss": 0.0451, "step": 20320 }, { "epoch": 0.4062587426561688, "grad_norm": 0.12341209501028061, "learning_rate": 9.838542958028516e-05, "loss": 0.0463, "step": 20330 }, { "epoch": 0.4064585747971704, "grad_norm": 0.15158399939537048, "learning_rate": 9.838282359214554e-05, "loss": 0.0428, "step": 20340 }, { "epoch": 0.40665840693817196, "grad_norm": 0.13271154463291168, "learning_rate": 9.83802155371806e-05, "loss": 0.0442, "step": 20350 }, { "epoch": 0.4068582390791735, "grad_norm": 0.15397202968597412, "learning_rate": 9.837760541550172e-05, "loss": 0.0451, "step": 20360 }, { "epoch": 0.40705807122017507, "grad_norm": 0.1501355916261673, "learning_rate": 9.837499322722039e-05, "loss": 0.0432, "step": 20370 }, { "epoch": 0.4072579033611766, "grad_norm": 0.12442188709974289, "learning_rate": 9.837237897244824e-05, "loss": 0.0476, "step": 20380 }, { "epoch": 0.4074577355021782, "grad_norm": 0.19037148356437683, "learning_rate": 9.836976265129691e-05, "loss": 0.0514, "step": 20390 }, { "epoch": 0.40765756764317973, "grad_norm": 0.16884486377239227, "learning_rate": 9.83671442638782e-05, "loss": 0.0418, "step": 20400 }, { "epoch": 0.4078573997841813, "grad_norm": 0.15327833592891693, "learning_rate": 9.836452381030392e-05, "loss": 0.0488, "step": 20410 }, { "epoch": 0.40805723192518284, "grad_norm": 0.1371428668498993, "learning_rate": 9.836190129068603e-05, "loss": 0.0443, "step": 20420 }, { "epoch": 0.4082570640661844, "grad_norm": 0.18913759291172028, "learning_rate": 9.835927670513657e-05, "loss": 0.0394, "step": 20430 }, { "epoch": 0.40845689620718595, "grad_norm": 0.15079401433467865, "learning_rate": 9.835665005376766e-05, "loss": 0.0475, "step": 20440 }, { "epoch": 0.4086567283481875, "grad_norm": 0.14342205226421356, "learning_rate": 9.835402133669147e-05, "loss": 0.0484, "step": 20450 }, { "epoch": 0.40885656048918906, "grad_norm": 0.14294637739658356, "learning_rate": 9.835139055402031e-05, "loss": 0.0443, "step": 20460 }, { "epoch": 0.4090563926301906, "grad_norm": 0.1361333727836609, "learning_rate": 9.83487577058666e-05, "loss": 0.04, "step": 20470 }, { "epoch": 0.4092562247711922, "grad_norm": 0.14325670897960663, "learning_rate": 9.834612279234276e-05, "loss": 0.0433, "step": 20480 }, { "epoch": 0.4094560569121938, "grad_norm": 0.19367454946041107, "learning_rate": 9.834348581356138e-05, "loss": 0.0456, "step": 20490 }, { "epoch": 0.40965588905319533, "grad_norm": 0.18368692696094513, "learning_rate": 9.834084676963509e-05, "loss": 0.0422, "step": 20500 }, { "epoch": 0.4098557211941969, "grad_norm": 0.15116943418979645, "learning_rate": 9.833820566067665e-05, "loss": 0.0373, "step": 20510 }, { "epoch": 0.41005555333519844, "grad_norm": 0.636072039604187, "learning_rate": 9.833556248679882e-05, "loss": 0.0502, "step": 20520 }, { "epoch": 0.4102553854762, "grad_norm": 0.12731589376926422, "learning_rate": 9.83329172481146e-05, "loss": 0.0421, "step": 20530 }, { "epoch": 0.41045521761720155, "grad_norm": 0.142592653632164, "learning_rate": 9.833026994473694e-05, "loss": 0.0471, "step": 20540 }, { "epoch": 0.4106550497582031, "grad_norm": 0.16888128221035004, "learning_rate": 9.832762057677891e-05, "loss": 0.0602, "step": 20550 }, { "epoch": 0.41085488189920466, "grad_norm": 0.17602676153182983, "learning_rate": 9.832496914435372e-05, "loss": 0.0526, "step": 20560 }, { "epoch": 0.4110547140402062, "grad_norm": 0.1419224590063095, "learning_rate": 9.832231564757462e-05, "loss": 0.0515, "step": 20570 }, { "epoch": 0.41125454618120777, "grad_norm": 0.1773114949464798, "learning_rate": 9.831966008655497e-05, "loss": 0.046, "step": 20580 }, { "epoch": 0.4114543783222093, "grad_norm": 0.153004989027977, "learning_rate": 9.831700246140821e-05, "loss": 0.0467, "step": 20590 }, { "epoch": 0.4116542104632109, "grad_norm": 0.13434262573719025, "learning_rate": 9.831434277224786e-05, "loss": 0.0409, "step": 20600 }, { "epoch": 0.4118540426042125, "grad_norm": 0.15044854581356049, "learning_rate": 9.831168101918755e-05, "loss": 0.0378, "step": 20610 }, { "epoch": 0.41205387474521404, "grad_norm": 0.1475576013326645, "learning_rate": 9.830901720234098e-05, "loss": 0.0508, "step": 20620 }, { "epoch": 0.4122537068862156, "grad_norm": 0.1736917942762375, "learning_rate": 9.830635132182194e-05, "loss": 0.0473, "step": 20630 }, { "epoch": 0.41245353902721715, "grad_norm": 0.14724309742450714, "learning_rate": 9.830368337774431e-05, "loss": 0.0437, "step": 20640 }, { "epoch": 0.4126533711682187, "grad_norm": 0.13138307631015778, "learning_rate": 9.830101337022206e-05, "loss": 0.0338, "step": 20650 }, { "epoch": 0.41285320330922026, "grad_norm": 0.18777890503406525, "learning_rate": 9.829834129936926e-05, "loss": 0.0485, "step": 20660 }, { "epoch": 0.4130530354502218, "grad_norm": 0.13260427117347717, "learning_rate": 9.829566716530004e-05, "loss": 0.0384, "step": 20670 }, { "epoch": 0.41325286759122337, "grad_norm": 0.14930443465709686, "learning_rate": 9.829299096812864e-05, "loss": 0.0408, "step": 20680 }, { "epoch": 0.4134526997322249, "grad_norm": 0.14619804918766022, "learning_rate": 9.829031270796938e-05, "loss": 0.0478, "step": 20690 }, { "epoch": 0.4136525318732265, "grad_norm": 0.1527135819196701, "learning_rate": 9.828763238493668e-05, "loss": 0.0391, "step": 20700 }, { "epoch": 0.41385236401422804, "grad_norm": 0.1779048889875412, "learning_rate": 9.828494999914503e-05, "loss": 0.0372, "step": 20710 }, { "epoch": 0.4140521961552296, "grad_norm": 0.14455202221870422, "learning_rate": 9.828226555070902e-05, "loss": 0.038, "step": 20720 }, { "epoch": 0.41425202829623115, "grad_norm": 0.1274242252111435, "learning_rate": 9.827957903974332e-05, "loss": 0.0512, "step": 20730 }, { "epoch": 0.4144518604372327, "grad_norm": 0.20320038497447968, "learning_rate": 9.827689046636271e-05, "loss": 0.0488, "step": 20740 }, { "epoch": 0.4146516925782343, "grad_norm": 0.1342175155878067, "learning_rate": 9.8274199830682e-05, "loss": 0.041, "step": 20750 }, { "epoch": 0.41485152471923586, "grad_norm": 0.15158316493034363, "learning_rate": 9.827150713281616e-05, "loss": 0.0514, "step": 20760 }, { "epoch": 0.4150513568602374, "grad_norm": 0.15285876393318176, "learning_rate": 9.826881237288023e-05, "loss": 0.0405, "step": 20770 }, { "epoch": 0.415251189001239, "grad_norm": 0.13810917735099792, "learning_rate": 9.82661155509893e-05, "loss": 0.0403, "step": 20780 }, { "epoch": 0.41545102114224053, "grad_norm": 0.12464381754398346, "learning_rate": 9.826341666725858e-05, "loss": 0.0526, "step": 20790 }, { "epoch": 0.4156508532832421, "grad_norm": 0.13778892159461975, "learning_rate": 9.826071572180337e-05, "loss": 0.0377, "step": 20800 }, { "epoch": 0.41585068542424364, "grad_norm": 0.13708704710006714, "learning_rate": 9.825801271473903e-05, "loss": 0.0464, "step": 20810 }, { "epoch": 0.4160505175652452, "grad_norm": 0.171994149684906, "learning_rate": 9.825530764618105e-05, "loss": 0.0396, "step": 20820 }, { "epoch": 0.41625034970624675, "grad_norm": 0.12297652661800385, "learning_rate": 9.825260051624496e-05, "loss": 0.0452, "step": 20830 }, { "epoch": 0.4164501818472483, "grad_norm": 0.14804722368717194, "learning_rate": 9.824989132504645e-05, "loss": 0.0396, "step": 20840 }, { "epoch": 0.41665001398824986, "grad_norm": 0.12289600074291229, "learning_rate": 9.82471800727012e-05, "loss": 0.0393, "step": 20850 }, { "epoch": 0.4168498461292514, "grad_norm": 0.1284864842891693, "learning_rate": 9.824446675932504e-05, "loss": 0.048, "step": 20860 }, { "epoch": 0.41704967827025297, "grad_norm": 0.12592124938964844, "learning_rate": 9.82417513850339e-05, "loss": 0.0478, "step": 20870 }, { "epoch": 0.4172495104112545, "grad_norm": 0.11780717223882675, "learning_rate": 9.823903394994376e-05, "loss": 0.0463, "step": 20880 }, { "epoch": 0.41744934255225613, "grad_norm": 0.14646993577480316, "learning_rate": 9.823631445417072e-05, "loss": 0.0438, "step": 20890 }, { "epoch": 0.4176491746932577, "grad_norm": 0.13785025477409363, "learning_rate": 9.823359289783092e-05, "loss": 0.0486, "step": 20900 }, { "epoch": 0.41784900683425924, "grad_norm": 0.12416892498731613, "learning_rate": 9.823086928104065e-05, "loss": 0.036, "step": 20910 }, { "epoch": 0.4180488389752608, "grad_norm": 0.14441610872745514, "learning_rate": 9.822814360391624e-05, "loss": 0.0479, "step": 20920 }, { "epoch": 0.41824867111626235, "grad_norm": 0.13504032790660858, "learning_rate": 9.822541586657414e-05, "loss": 0.05, "step": 20930 }, { "epoch": 0.4184485032572639, "grad_norm": 0.14681193232536316, "learning_rate": 9.822268606913088e-05, "loss": 0.039, "step": 20940 }, { "epoch": 0.41864833539826546, "grad_norm": 0.19005727767944336, "learning_rate": 9.821995421170304e-05, "loss": 0.0464, "step": 20950 }, { "epoch": 0.418848167539267, "grad_norm": 0.1519973874092102, "learning_rate": 9.821722029440735e-05, "loss": 0.0419, "step": 20960 }, { "epoch": 0.41904799968026857, "grad_norm": 0.21523739397525787, "learning_rate": 9.821448431736058e-05, "loss": 0.0408, "step": 20970 }, { "epoch": 0.4192478318212701, "grad_norm": 0.16306708753108978, "learning_rate": 9.82117462806796e-05, "loss": 0.0417, "step": 20980 }, { "epoch": 0.4194476639622717, "grad_norm": 0.15427164733409882, "learning_rate": 9.820900618448142e-05, "loss": 0.0445, "step": 20990 }, { "epoch": 0.41964749610327323, "grad_norm": 0.12855368852615356, "learning_rate": 9.820626402888303e-05, "loss": 0.0411, "step": 21000 }, { "epoch": 0.4198473282442748, "grad_norm": 0.12714633345603943, "learning_rate": 9.82035198140016e-05, "loss": 0.0422, "step": 21010 }, { "epoch": 0.42004716038527634, "grad_norm": 0.19735391438007355, "learning_rate": 9.820077353995436e-05, "loss": 0.0533, "step": 21020 }, { "epoch": 0.42024699252627795, "grad_norm": 0.1549273282289505, "learning_rate": 9.819802520685864e-05, "loss": 0.0543, "step": 21030 }, { "epoch": 0.4204468246672795, "grad_norm": 0.165220245718956, "learning_rate": 9.819527481483181e-05, "loss": 0.0458, "step": 21040 }, { "epoch": 0.42064665680828106, "grad_norm": 0.15987884998321533, "learning_rate": 9.819252236399137e-05, "loss": 0.05, "step": 21050 }, { "epoch": 0.4208464889492826, "grad_norm": 0.2168273627758026, "learning_rate": 9.818976785445492e-05, "loss": 0.0478, "step": 21060 }, { "epoch": 0.42104632109028417, "grad_norm": 0.1539095789194107, "learning_rate": 9.818701128634011e-05, "loss": 0.0408, "step": 21070 }, { "epoch": 0.4212461532312857, "grad_norm": 0.1321316808462143, "learning_rate": 9.81842526597647e-05, "loss": 0.046, "step": 21080 }, { "epoch": 0.4214459853722873, "grad_norm": 0.1490178406238556, "learning_rate": 9.818149197484654e-05, "loss": 0.038, "step": 21090 }, { "epoch": 0.42164581751328883, "grad_norm": 0.13870316743850708, "learning_rate": 9.817872923170355e-05, "loss": 0.0412, "step": 21100 }, { "epoch": 0.4218456496542904, "grad_norm": 0.14033083617687225, "learning_rate": 9.817596443045375e-05, "loss": 0.0419, "step": 21110 }, { "epoch": 0.42204548179529194, "grad_norm": 0.16697721183300018, "learning_rate": 9.817319757121526e-05, "loss": 0.0474, "step": 21120 }, { "epoch": 0.4222453139362935, "grad_norm": 0.15863585472106934, "learning_rate": 9.817042865410626e-05, "loss": 0.0456, "step": 21130 }, { "epoch": 0.42244514607729505, "grad_norm": 0.16790160536766052, "learning_rate": 9.816765767924505e-05, "loss": 0.0468, "step": 21140 }, { "epoch": 0.4226449782182966, "grad_norm": 0.13031725585460663, "learning_rate": 9.816488464674998e-05, "loss": 0.0387, "step": 21150 }, { "epoch": 0.42284481035929816, "grad_norm": 0.16515296697616577, "learning_rate": 9.816210955673952e-05, "loss": 0.043, "step": 21160 }, { "epoch": 0.4230446425002998, "grad_norm": 0.1831057071685791, "learning_rate": 9.815933240933222e-05, "loss": 0.0483, "step": 21170 }, { "epoch": 0.4232444746413013, "grad_norm": 0.14799463748931885, "learning_rate": 9.815655320464671e-05, "loss": 0.0476, "step": 21180 }, { "epoch": 0.4234443067823029, "grad_norm": 0.14573128521442413, "learning_rate": 9.815377194280171e-05, "loss": 0.0419, "step": 21190 }, { "epoch": 0.42364413892330444, "grad_norm": 0.12977395951747894, "learning_rate": 9.815098862391604e-05, "loss": 0.0431, "step": 21200 }, { "epoch": 0.423843971064306, "grad_norm": 0.1464255154132843, "learning_rate": 9.814820324810858e-05, "loss": 0.0405, "step": 21210 }, { "epoch": 0.42404380320530755, "grad_norm": 0.1740129292011261, "learning_rate": 9.814541581549834e-05, "loss": 0.0466, "step": 21220 }, { "epoch": 0.4242436353463091, "grad_norm": 0.1524435132741928, "learning_rate": 9.814262632620438e-05, "loss": 0.047, "step": 21230 }, { "epoch": 0.42444346748731065, "grad_norm": 0.11640413105487823, "learning_rate": 9.813983478034587e-05, "loss": 0.0374, "step": 21240 }, { "epoch": 0.4246432996283122, "grad_norm": 0.18945004045963287, "learning_rate": 9.813704117804205e-05, "loss": 0.0439, "step": 21250 }, { "epoch": 0.42484313176931376, "grad_norm": 0.12562765181064606, "learning_rate": 9.813424551941226e-05, "loss": 0.0476, "step": 21260 }, { "epoch": 0.4250429639103153, "grad_norm": 0.14180083572864532, "learning_rate": 9.813144780457593e-05, "loss": 0.0447, "step": 21270 }, { "epoch": 0.4252427960513169, "grad_norm": 0.13234038650989532, "learning_rate": 9.812864803365256e-05, "loss": 0.042, "step": 21280 }, { "epoch": 0.42544262819231843, "grad_norm": 0.159496009349823, "learning_rate": 9.812584620676178e-05, "loss": 0.0353, "step": 21290 }, { "epoch": 0.42564246033332004, "grad_norm": 0.17738163471221924, "learning_rate": 9.812304232402324e-05, "loss": 0.0428, "step": 21300 }, { "epoch": 0.4258422924743216, "grad_norm": 0.14933991432189941, "learning_rate": 9.812023638555673e-05, "loss": 0.043, "step": 21310 }, { "epoch": 0.42604212461532315, "grad_norm": 0.10690375417470932, "learning_rate": 9.811742839148214e-05, "loss": 0.0442, "step": 21320 }, { "epoch": 0.4262419567563247, "grad_norm": 0.1563752442598343, "learning_rate": 9.81146183419194e-05, "loss": 0.0437, "step": 21330 }, { "epoch": 0.42644178889732626, "grad_norm": 0.15603791177272797, "learning_rate": 9.811180623698857e-05, "loss": 0.04, "step": 21340 }, { "epoch": 0.4266416210383278, "grad_norm": 0.12149938195943832, "learning_rate": 9.810899207680973e-05, "loss": 0.0418, "step": 21350 }, { "epoch": 0.42684145317932937, "grad_norm": 0.16558386385440826, "learning_rate": 9.810617586150315e-05, "loss": 0.0454, "step": 21360 }, { "epoch": 0.4270412853203309, "grad_norm": 0.16690665483474731, "learning_rate": 9.81033575911891e-05, "loss": 0.039, "step": 21370 }, { "epoch": 0.4272411174613325, "grad_norm": 0.14687317609786987, "learning_rate": 9.810053726598799e-05, "loss": 0.0454, "step": 21380 }, { "epoch": 0.42744094960233403, "grad_norm": 0.1988137662410736, "learning_rate": 9.809771488602029e-05, "loss": 0.0424, "step": 21390 }, { "epoch": 0.4276407817433356, "grad_norm": 0.12506292760372162, "learning_rate": 9.809489045140657e-05, "loss": 0.0454, "step": 21400 }, { "epoch": 0.42784061388433714, "grad_norm": 0.14610828459262848, "learning_rate": 9.809206396226747e-05, "loss": 0.0415, "step": 21410 }, { "epoch": 0.4280404460253387, "grad_norm": 0.1417626589536667, "learning_rate": 9.808923541872376e-05, "loss": 0.056, "step": 21420 }, { "epoch": 0.42824027816634025, "grad_norm": 0.20317062735557556, "learning_rate": 9.808640482089626e-05, "loss": 0.0388, "step": 21430 }, { "epoch": 0.42844011030734186, "grad_norm": 0.10886791348457336, "learning_rate": 9.808357216890587e-05, "loss": 0.0413, "step": 21440 }, { "epoch": 0.4286399424483434, "grad_norm": 0.12871691584587097, "learning_rate": 9.80807374628736e-05, "loss": 0.0458, "step": 21450 }, { "epoch": 0.42883977458934497, "grad_norm": 0.11601460725069046, "learning_rate": 9.807790070292058e-05, "loss": 0.0343, "step": 21460 }, { "epoch": 0.4290396067303465, "grad_norm": 0.16144098341464996, "learning_rate": 9.807506188916794e-05, "loss": 0.0468, "step": 21470 }, { "epoch": 0.4292394388713481, "grad_norm": 0.18615110218524933, "learning_rate": 9.807222102173698e-05, "loss": 0.0461, "step": 21480 }, { "epoch": 0.42943927101234963, "grad_norm": 0.12896133959293365, "learning_rate": 9.806937810074906e-05, "loss": 0.0425, "step": 21490 }, { "epoch": 0.4296391031533512, "grad_norm": 0.18997876346111298, "learning_rate": 9.80665331263256e-05, "loss": 0.042, "step": 21500 }, { "epoch": 0.42983893529435274, "grad_norm": 0.17241205275058746, "learning_rate": 9.806368609858815e-05, "loss": 0.041, "step": 21510 }, { "epoch": 0.4300387674353543, "grad_norm": 0.1494637131690979, "learning_rate": 9.806083701765834e-05, "loss": 0.0506, "step": 21520 }, { "epoch": 0.43023859957635585, "grad_norm": 0.10807838290929794, "learning_rate": 9.805798588365785e-05, "loss": 0.0428, "step": 21530 }, { "epoch": 0.4304384317173574, "grad_norm": 0.13909542560577393, "learning_rate": 9.805513269670848e-05, "loss": 0.0463, "step": 21540 }, { "epoch": 0.43063826385835896, "grad_norm": 0.1465802788734436, "learning_rate": 9.805227745693213e-05, "loss": 0.0448, "step": 21550 }, { "epoch": 0.4308380959993605, "grad_norm": 0.12964767217636108, "learning_rate": 9.804942016445075e-05, "loss": 0.0429, "step": 21560 }, { "epoch": 0.43103792814036207, "grad_norm": 0.10954742133617401, "learning_rate": 9.804656081938643e-05, "loss": 0.0443, "step": 21570 }, { "epoch": 0.4312377602813637, "grad_norm": 0.14991597831249237, "learning_rate": 9.804369942186128e-05, "loss": 0.0453, "step": 21580 }, { "epoch": 0.43143759242236523, "grad_norm": 0.12041779607534409, "learning_rate": 9.804083597199758e-05, "loss": 0.0469, "step": 21590 }, { "epoch": 0.4316374245633668, "grad_norm": 0.15000580251216888, "learning_rate": 9.80379704699176e-05, "loss": 0.0408, "step": 21600 }, { "epoch": 0.43183725670436834, "grad_norm": 0.14152315258979797, "learning_rate": 9.803510291574376e-05, "loss": 0.039, "step": 21610 }, { "epoch": 0.4320370888453699, "grad_norm": 0.18202628195285797, "learning_rate": 9.803223330959858e-05, "loss": 0.0463, "step": 21620 }, { "epoch": 0.43223692098637145, "grad_norm": 0.14654864370822906, "learning_rate": 9.802936165160464e-05, "loss": 0.0545, "step": 21630 }, { "epoch": 0.432436753127373, "grad_norm": 0.182673841714859, "learning_rate": 9.802648794188458e-05, "loss": 0.0444, "step": 21640 }, { "epoch": 0.43263658526837456, "grad_norm": 0.17517337203025818, "learning_rate": 9.80236121805612e-05, "loss": 0.0467, "step": 21650 }, { "epoch": 0.4328364174093761, "grad_norm": 0.14623674750328064, "learning_rate": 9.802073436775732e-05, "loss": 0.051, "step": 21660 }, { "epoch": 0.43303624955037767, "grad_norm": 0.1808248907327652, "learning_rate": 9.801785450359589e-05, "loss": 0.0483, "step": 21670 }, { "epoch": 0.4332360816913792, "grad_norm": 0.14343766868114471, "learning_rate": 9.801497258819992e-05, "loss": 0.0419, "step": 21680 }, { "epoch": 0.4334359138323808, "grad_norm": 0.14727304875850677, "learning_rate": 9.801208862169253e-05, "loss": 0.0464, "step": 21690 }, { "epoch": 0.43363574597338234, "grad_norm": 0.1969672292470932, "learning_rate": 9.800920260419692e-05, "loss": 0.0378, "step": 21700 }, { "epoch": 0.4338355781143839, "grad_norm": 0.12833815813064575, "learning_rate": 9.800631453583636e-05, "loss": 0.0368, "step": 21710 }, { "epoch": 0.4340354102553855, "grad_norm": 0.10961606353521347, "learning_rate": 9.800342441673425e-05, "loss": 0.0459, "step": 21720 }, { "epoch": 0.43423524239638706, "grad_norm": 0.17042036354541779, "learning_rate": 9.800053224701402e-05, "loss": 0.0448, "step": 21730 }, { "epoch": 0.4344350745373886, "grad_norm": 0.12772807478904724, "learning_rate": 9.799763802679924e-05, "loss": 0.0382, "step": 21740 }, { "epoch": 0.43463490667839016, "grad_norm": 0.15408235788345337, "learning_rate": 9.799474175621352e-05, "loss": 0.0417, "step": 21750 }, { "epoch": 0.4348347388193917, "grad_norm": 0.15226076543331146, "learning_rate": 9.799184343538062e-05, "loss": 0.0425, "step": 21760 }, { "epoch": 0.4350345709603933, "grad_norm": 0.17116330564022064, "learning_rate": 9.798894306442435e-05, "loss": 0.0413, "step": 21770 }, { "epoch": 0.43523440310139483, "grad_norm": 0.1479143351316452, "learning_rate": 9.798604064346855e-05, "loss": 0.044, "step": 21780 }, { "epoch": 0.4354342352423964, "grad_norm": 0.15247565507888794, "learning_rate": 9.798313617263728e-05, "loss": 0.0482, "step": 21790 }, { "epoch": 0.43563406738339794, "grad_norm": 0.13408994674682617, "learning_rate": 9.798022965205456e-05, "loss": 0.0333, "step": 21800 }, { "epoch": 0.4358338995243995, "grad_norm": 0.16002927720546722, "learning_rate": 9.797732108184458e-05, "loss": 0.0451, "step": 21810 }, { "epoch": 0.43603373166540105, "grad_norm": 0.14201249182224274, "learning_rate": 9.797441046213158e-05, "loss": 0.0409, "step": 21820 }, { "epoch": 0.4362335638064026, "grad_norm": 0.1321788877248764, "learning_rate": 9.797149779303989e-05, "loss": 0.0364, "step": 21830 }, { "epoch": 0.43643339594740416, "grad_norm": 0.14162011444568634, "learning_rate": 9.796858307469395e-05, "loss": 0.0409, "step": 21840 }, { "epoch": 0.4366332280884057, "grad_norm": 0.12877878546714783, "learning_rate": 9.796566630721826e-05, "loss": 0.0514, "step": 21850 }, { "epoch": 0.4368330602294073, "grad_norm": 0.1462578922510147, "learning_rate": 9.796274749073742e-05, "loss": 0.0441, "step": 21860 }, { "epoch": 0.4370328923704089, "grad_norm": 0.14974910020828247, "learning_rate": 9.795982662537613e-05, "loss": 0.0382, "step": 21870 }, { "epoch": 0.43723272451141043, "grad_norm": 0.1345260590314865, "learning_rate": 9.795690371125913e-05, "loss": 0.0438, "step": 21880 }, { "epoch": 0.437432556652412, "grad_norm": 0.16047658026218414, "learning_rate": 9.79539787485113e-05, "loss": 0.0514, "step": 21890 }, { "epoch": 0.43763238879341354, "grad_norm": 0.1347947120666504, "learning_rate": 9.795105173725761e-05, "loss": 0.04, "step": 21900 }, { "epoch": 0.4378322209344151, "grad_norm": 0.10902785509824753, "learning_rate": 9.794812267762308e-05, "loss": 0.0372, "step": 21910 }, { "epoch": 0.43803205307541665, "grad_norm": 0.1715276539325714, "learning_rate": 9.794519156973283e-05, "loss": 0.0448, "step": 21920 }, { "epoch": 0.4382318852164182, "grad_norm": 0.15889972448349, "learning_rate": 9.794225841371208e-05, "loss": 0.0426, "step": 21930 }, { "epoch": 0.43843171735741976, "grad_norm": 0.17753304541110992, "learning_rate": 9.79393232096861e-05, "loss": 0.0438, "step": 21940 }, { "epoch": 0.4386315494984213, "grad_norm": 0.14181339740753174, "learning_rate": 9.793638595778032e-05, "loss": 0.0481, "step": 21950 }, { "epoch": 0.43883138163942287, "grad_norm": 0.16323615610599518, "learning_rate": 9.793344665812018e-05, "loss": 0.042, "step": 21960 }, { "epoch": 0.4390312137804244, "grad_norm": 0.1282547563314438, "learning_rate": 9.793050531083127e-05, "loss": 0.0456, "step": 21970 }, { "epoch": 0.439231045921426, "grad_norm": 0.1659250408411026, "learning_rate": 9.792756191603921e-05, "loss": 0.0498, "step": 21980 }, { "epoch": 0.4394308780624276, "grad_norm": 0.16189447045326233, "learning_rate": 9.792461647386976e-05, "loss": 0.0496, "step": 21990 }, { "epoch": 0.43963071020342914, "grad_norm": 0.12658509612083435, "learning_rate": 9.792166898444873e-05, "loss": 0.0542, "step": 22000 }, { "epoch": 0.4398305423444307, "grad_norm": 0.14175337553024292, "learning_rate": 9.791871944790204e-05, "loss": 0.0421, "step": 22010 }, { "epoch": 0.44003037448543225, "grad_norm": 0.12463077902793884, "learning_rate": 9.791576786435568e-05, "loss": 0.0378, "step": 22020 }, { "epoch": 0.4402302066264338, "grad_norm": 0.14119133353233337, "learning_rate": 9.791281423393575e-05, "loss": 0.0437, "step": 22030 }, { "epoch": 0.44043003876743536, "grad_norm": 0.18929824233055115, "learning_rate": 9.79098585567684e-05, "loss": 0.0442, "step": 22040 }, { "epoch": 0.4406298709084369, "grad_norm": 0.21308675408363342, "learning_rate": 9.790690083297991e-05, "loss": 0.059, "step": 22050 }, { "epoch": 0.44082970304943847, "grad_norm": 0.16835863888263702, "learning_rate": 9.790394106269664e-05, "loss": 0.0434, "step": 22060 }, { "epoch": 0.44102953519044, "grad_norm": 0.16491656005382538, "learning_rate": 9.7900979246045e-05, "loss": 0.039, "step": 22070 }, { "epoch": 0.4412293673314416, "grad_norm": 0.15616051852703094, "learning_rate": 9.789801538315152e-05, "loss": 0.0398, "step": 22080 }, { "epoch": 0.44142919947244313, "grad_norm": 0.15240192413330078, "learning_rate": 9.789504947414281e-05, "loss": 0.0378, "step": 22090 }, { "epoch": 0.4416290316134447, "grad_norm": 0.1796259731054306, "learning_rate": 9.789208151914559e-05, "loss": 0.051, "step": 22100 }, { "epoch": 0.44182886375444624, "grad_norm": 0.1337026208639145, "learning_rate": 9.788911151828661e-05, "loss": 0.0372, "step": 22110 }, { "epoch": 0.4420286958954478, "grad_norm": 0.12074685096740723, "learning_rate": 9.788613947169276e-05, "loss": 0.0362, "step": 22120 }, { "epoch": 0.4422285280364494, "grad_norm": 0.16374485194683075, "learning_rate": 9.788316537949101e-05, "loss": 0.0522, "step": 22130 }, { "epoch": 0.44242836017745096, "grad_norm": 0.12773080170154572, "learning_rate": 9.788018924180841e-05, "loss": 0.0458, "step": 22140 }, { "epoch": 0.4426281923184525, "grad_norm": 0.17073982954025269, "learning_rate": 9.787721105877206e-05, "loss": 0.0476, "step": 22150 }, { "epoch": 0.4428280244594541, "grad_norm": 0.1277562379837036, "learning_rate": 9.787423083050923e-05, "loss": 0.0457, "step": 22160 }, { "epoch": 0.4430278566004556, "grad_norm": 0.1846310943365097, "learning_rate": 9.78712485571472e-05, "loss": 0.0469, "step": 22170 }, { "epoch": 0.4432276887414572, "grad_norm": 0.13425791263580322, "learning_rate": 9.786826423881337e-05, "loss": 0.041, "step": 22180 }, { "epoch": 0.44342752088245874, "grad_norm": 0.12889713048934937, "learning_rate": 9.786527787563522e-05, "loss": 0.0446, "step": 22190 }, { "epoch": 0.4436273530234603, "grad_norm": 0.20257331430912018, "learning_rate": 9.786228946774034e-05, "loss": 0.0462, "step": 22200 }, { "epoch": 0.44382718516446185, "grad_norm": 0.18804781138896942, "learning_rate": 9.785929901525639e-05, "loss": 0.0589, "step": 22210 }, { "epoch": 0.4440270173054634, "grad_norm": 0.13764464855194092, "learning_rate": 9.785630651831108e-05, "loss": 0.0448, "step": 22220 }, { "epoch": 0.44422684944646496, "grad_norm": 0.159162238240242, "learning_rate": 9.785331197703229e-05, "loss": 0.0449, "step": 22230 }, { "epoch": 0.4444266815874665, "grad_norm": 0.15737171471118927, "learning_rate": 9.785031539154791e-05, "loss": 0.0478, "step": 22240 }, { "epoch": 0.44462651372846806, "grad_norm": 0.18463799357414246, "learning_rate": 9.784731676198597e-05, "loss": 0.041, "step": 22250 }, { "epoch": 0.4448263458694696, "grad_norm": 0.15072932839393616, "learning_rate": 9.784431608847456e-05, "loss": 0.0453, "step": 22260 }, { "epoch": 0.44502617801047123, "grad_norm": 0.14452242851257324, "learning_rate": 9.784131337114186e-05, "loss": 0.0464, "step": 22270 }, { "epoch": 0.4452260101514728, "grad_norm": 0.14304383099079132, "learning_rate": 9.783830861011613e-05, "loss": 0.0505, "step": 22280 }, { "epoch": 0.44542584229247434, "grad_norm": 0.14827117323875427, "learning_rate": 9.783530180552574e-05, "loss": 0.0571, "step": 22290 }, { "epoch": 0.4456256744334759, "grad_norm": 0.11870969086885452, "learning_rate": 9.783229295749914e-05, "loss": 0.0436, "step": 22300 }, { "epoch": 0.44582550657447745, "grad_norm": 0.13659721612930298, "learning_rate": 9.782928206616485e-05, "loss": 0.0465, "step": 22310 }, { "epoch": 0.446025338715479, "grad_norm": 0.12746047973632812, "learning_rate": 9.782626913165149e-05, "loss": 0.05, "step": 22320 }, { "epoch": 0.44622517085648056, "grad_norm": 0.18446560204029083, "learning_rate": 9.782325415408778e-05, "loss": 0.0455, "step": 22330 }, { "epoch": 0.4464250029974821, "grad_norm": 0.15723134577274323, "learning_rate": 9.78202371336025e-05, "loss": 0.0518, "step": 22340 }, { "epoch": 0.44662483513848367, "grad_norm": 0.13479316234588623, "learning_rate": 9.781721807032455e-05, "loss": 0.0515, "step": 22350 }, { "epoch": 0.4468246672794852, "grad_norm": 0.18595640361309052, "learning_rate": 9.781419696438288e-05, "loss": 0.0369, "step": 22360 }, { "epoch": 0.4470244994204868, "grad_norm": 0.15106983482837677, "learning_rate": 9.781117381590655e-05, "loss": 0.0452, "step": 22370 }, { "epoch": 0.44722433156148833, "grad_norm": 0.1693711131811142, "learning_rate": 9.78081486250247e-05, "loss": 0.0539, "step": 22380 }, { "epoch": 0.4474241637024899, "grad_norm": 0.18100249767303467, "learning_rate": 9.78051213918666e-05, "loss": 0.0499, "step": 22390 }, { "epoch": 0.44762399584349144, "grad_norm": 0.18797147274017334, "learning_rate": 9.78020921165615e-05, "loss": 0.0448, "step": 22400 }, { "epoch": 0.44782382798449305, "grad_norm": 0.1353788524866104, "learning_rate": 9.779906079923886e-05, "loss": 0.0519, "step": 22410 }, { "epoch": 0.4480236601254946, "grad_norm": 0.1836465448141098, "learning_rate": 9.779602744002814e-05, "loss": 0.0487, "step": 22420 }, { "epoch": 0.44822349226649616, "grad_norm": 0.16640962660312653, "learning_rate": 9.779299203905893e-05, "loss": 0.0356, "step": 22430 }, { "epoch": 0.4484233244074977, "grad_norm": 0.1487727165222168, "learning_rate": 9.77899545964609e-05, "loss": 0.0497, "step": 22440 }, { "epoch": 0.44862315654849927, "grad_norm": 0.13313820958137512, "learning_rate": 9.778691511236382e-05, "loss": 0.0444, "step": 22450 }, { "epoch": 0.4488229886895008, "grad_norm": 0.19235865771770477, "learning_rate": 9.778387358689749e-05, "loss": 0.0525, "step": 22460 }, { "epoch": 0.4490228208305024, "grad_norm": 0.17665810883045197, "learning_rate": 9.778083002019188e-05, "loss": 0.0438, "step": 22470 }, { "epoch": 0.44922265297150393, "grad_norm": 0.14879858493804932, "learning_rate": 9.777778441237697e-05, "loss": 0.0386, "step": 22480 }, { "epoch": 0.4494224851125055, "grad_norm": 0.12339390814304352, "learning_rate": 9.77747367635829e-05, "loss": 0.0556, "step": 22490 }, { "epoch": 0.44962231725350704, "grad_norm": 0.13409434258937836, "learning_rate": 9.777168707393983e-05, "loss": 0.0452, "step": 22500 }, { "epoch": 0.4498221493945086, "grad_norm": 0.18555691838264465, "learning_rate": 9.776863534357806e-05, "loss": 0.0418, "step": 22510 }, { "epoch": 0.45002198153551015, "grad_norm": 0.15317267179489136, "learning_rate": 9.776558157262792e-05, "loss": 0.0383, "step": 22520 }, { "epoch": 0.4502218136765117, "grad_norm": 0.1366288810968399, "learning_rate": 9.77625257612199e-05, "loss": 0.044, "step": 22530 }, { "epoch": 0.4504216458175133, "grad_norm": 0.16683360934257507, "learning_rate": 9.775946790948451e-05, "loss": 0.04, "step": 22540 }, { "epoch": 0.45062147795851487, "grad_norm": 0.14411978423595428, "learning_rate": 9.775640801755239e-05, "loss": 0.0417, "step": 22550 }, { "epoch": 0.4508213100995164, "grad_norm": 0.1842058151960373, "learning_rate": 9.775334608555427e-05, "loss": 0.0457, "step": 22560 }, { "epoch": 0.451021142240518, "grad_norm": 0.13476578891277313, "learning_rate": 9.775028211362091e-05, "loss": 0.0425, "step": 22570 }, { "epoch": 0.45122097438151954, "grad_norm": 0.18445216119289398, "learning_rate": 9.774721610188323e-05, "loss": 0.0462, "step": 22580 }, { "epoch": 0.4514208065225211, "grad_norm": 0.15935032069683075, "learning_rate": 9.774414805047218e-05, "loss": 0.0478, "step": 22590 }, { "epoch": 0.45162063866352264, "grad_norm": 0.1506585031747818, "learning_rate": 9.774107795951883e-05, "loss": 0.0435, "step": 22600 }, { "epoch": 0.4518204708045242, "grad_norm": 0.12215229868888855, "learning_rate": 9.773800582915436e-05, "loss": 0.0353, "step": 22610 }, { "epoch": 0.45202030294552575, "grad_norm": 0.11299595981836319, "learning_rate": 9.773493165950995e-05, "loss": 0.0454, "step": 22620 }, { "epoch": 0.4522201350865273, "grad_norm": 0.17023056745529175, "learning_rate": 9.773185545071697e-05, "loss": 0.0401, "step": 22630 }, { "epoch": 0.45241996722752886, "grad_norm": 0.16019465029239655, "learning_rate": 9.77287772029068e-05, "loss": 0.0417, "step": 22640 }, { "epoch": 0.4526197993685304, "grad_norm": 0.13921262323856354, "learning_rate": 9.772569691621097e-05, "loss": 0.048, "step": 22650 }, { "epoch": 0.452819631509532, "grad_norm": 0.1164652481675148, "learning_rate": 9.772261459076103e-05, "loss": 0.0437, "step": 22660 }, { "epoch": 0.4530194636505335, "grad_norm": 0.13830454647541046, "learning_rate": 9.771953022668865e-05, "loss": 0.0448, "step": 22670 }, { "epoch": 0.45321929579153514, "grad_norm": 0.16200630366802216, "learning_rate": 9.771644382412561e-05, "loss": 0.055, "step": 22680 }, { "epoch": 0.4534191279325367, "grad_norm": 0.132231205701828, "learning_rate": 9.771335538320375e-05, "loss": 0.0402, "step": 22690 }, { "epoch": 0.45361896007353825, "grad_norm": 0.14065346121788025, "learning_rate": 9.771026490405499e-05, "loss": 0.0425, "step": 22700 }, { "epoch": 0.4538187922145398, "grad_norm": 0.15503115952014923, "learning_rate": 9.770717238681137e-05, "loss": 0.0407, "step": 22710 }, { "epoch": 0.45401862435554136, "grad_norm": 0.13190095126628876, "learning_rate": 9.7704077831605e-05, "loss": 0.0404, "step": 22720 }, { "epoch": 0.4542184564965429, "grad_norm": 0.15254028141498566, "learning_rate": 9.770098123856804e-05, "loss": 0.0497, "step": 22730 }, { "epoch": 0.45441828863754447, "grad_norm": 0.14749214053153992, "learning_rate": 9.76978826078328e-05, "loss": 0.0491, "step": 22740 }, { "epoch": 0.454618120778546, "grad_norm": 0.12307921051979065, "learning_rate": 9.769478193953163e-05, "loss": 0.0401, "step": 22750 }, { "epoch": 0.4548179529195476, "grad_norm": 0.13058188557624817, "learning_rate": 9.769167923379701e-05, "loss": 0.0414, "step": 22760 }, { "epoch": 0.45501778506054913, "grad_norm": 0.1315533071756363, "learning_rate": 9.768857449076146e-05, "loss": 0.0336, "step": 22770 }, { "epoch": 0.4552176172015507, "grad_norm": 0.13240110874176025, "learning_rate": 9.76854677105576e-05, "loss": 0.0486, "step": 22780 }, { "epoch": 0.45541744934255224, "grad_norm": 0.1498965471982956, "learning_rate": 9.768235889331819e-05, "loss": 0.0412, "step": 22790 }, { "epoch": 0.4556172814835538, "grad_norm": 0.14396710693836212, "learning_rate": 9.767924803917596e-05, "loss": 0.0356, "step": 22800 }, { "epoch": 0.45581711362455535, "grad_norm": 0.15713509917259216, "learning_rate": 9.767613514826388e-05, "loss": 0.0505, "step": 22810 }, { "epoch": 0.45601694576555696, "grad_norm": 0.18656277656555176, "learning_rate": 9.767302022071487e-05, "loss": 0.042, "step": 22820 }, { "epoch": 0.4562167779065585, "grad_norm": 0.16345401108264923, "learning_rate": 9.766990325666203e-05, "loss": 0.0421, "step": 22830 }, { "epoch": 0.45641661004756007, "grad_norm": 0.15417835116386414, "learning_rate": 9.766678425623849e-05, "loss": 0.0499, "step": 22840 }, { "epoch": 0.4566164421885616, "grad_norm": 0.13747188448905945, "learning_rate": 9.766366321957749e-05, "loss": 0.0408, "step": 22850 }, { "epoch": 0.4568162743295632, "grad_norm": 0.1382443755865097, "learning_rate": 9.766054014681237e-05, "loss": 0.0439, "step": 22860 }, { "epoch": 0.45701610647056473, "grad_norm": 0.14955106377601624, "learning_rate": 9.76574150380765e-05, "loss": 0.0386, "step": 22870 }, { "epoch": 0.4572159386115663, "grad_norm": 0.15202027559280396, "learning_rate": 9.765428789350344e-05, "loss": 0.0446, "step": 22880 }, { "epoch": 0.45741577075256784, "grad_norm": 0.16985191404819489, "learning_rate": 9.765115871322672e-05, "loss": 0.0406, "step": 22890 }, { "epoch": 0.4576156028935694, "grad_norm": 0.13073572516441345, "learning_rate": 9.764802749738006e-05, "loss": 0.0494, "step": 22900 }, { "epoch": 0.45781543503457095, "grad_norm": 0.1570989191532135, "learning_rate": 9.764489424609719e-05, "loss": 0.0476, "step": 22910 }, { "epoch": 0.4580152671755725, "grad_norm": 0.13388709723949432, "learning_rate": 9.764175895951196e-05, "loss": 0.0613, "step": 22920 }, { "epoch": 0.45821509931657406, "grad_norm": 0.15011484920978546, "learning_rate": 9.763862163775831e-05, "loss": 0.0474, "step": 22930 }, { "epoch": 0.4584149314575756, "grad_norm": 0.17585906386375427, "learning_rate": 9.763548228097026e-05, "loss": 0.0451, "step": 22940 }, { "epoch": 0.45861476359857717, "grad_norm": 0.11754598468542099, "learning_rate": 9.763234088928192e-05, "loss": 0.0447, "step": 22950 }, { "epoch": 0.4588145957395788, "grad_norm": 0.12980784475803375, "learning_rate": 9.762919746282747e-05, "loss": 0.0389, "step": 22960 }, { "epoch": 0.45901442788058033, "grad_norm": 0.14401713013648987, "learning_rate": 9.762605200174121e-05, "loss": 0.0327, "step": 22970 }, { "epoch": 0.4592142600215819, "grad_norm": 0.12183794379234314, "learning_rate": 9.76229045061575e-05, "loss": 0.0409, "step": 22980 }, { "epoch": 0.45941409216258344, "grad_norm": 0.12771350145339966, "learning_rate": 9.761975497621078e-05, "loss": 0.0406, "step": 22990 }, { "epoch": 0.459613924303585, "grad_norm": 0.12158400565385818, "learning_rate": 9.761660341203562e-05, "loss": 0.0468, "step": 23000 }, { "epoch": 0.45981375644458655, "grad_norm": 0.15653298795223236, "learning_rate": 9.761344981376665e-05, "loss": 0.0385, "step": 23010 }, { "epoch": 0.4600135885855881, "grad_norm": 0.1532670259475708, "learning_rate": 9.761029418153855e-05, "loss": 0.045, "step": 23020 }, { "epoch": 0.46021342072658966, "grad_norm": 0.14430813491344452, "learning_rate": 9.760713651548616e-05, "loss": 0.0376, "step": 23030 }, { "epoch": 0.4604132528675912, "grad_norm": 0.14087840914726257, "learning_rate": 9.760397681574436e-05, "loss": 0.0409, "step": 23040 }, { "epoch": 0.46061308500859277, "grad_norm": 0.10990644991397858, "learning_rate": 9.760081508244811e-05, "loss": 0.0454, "step": 23050 }, { "epoch": 0.4608129171495943, "grad_norm": 0.13571546971797943, "learning_rate": 9.759765131573249e-05, "loss": 0.0436, "step": 23060 }, { "epoch": 0.4610127492905959, "grad_norm": 0.19189868867397308, "learning_rate": 9.759448551573265e-05, "loss": 0.0399, "step": 23070 }, { "epoch": 0.46121258143159743, "grad_norm": 0.14756906032562256, "learning_rate": 9.75913176825838e-05, "loss": 0.0413, "step": 23080 }, { "epoch": 0.461412413572599, "grad_norm": 0.13713763654232025, "learning_rate": 9.758814781642132e-05, "loss": 0.0498, "step": 23090 }, { "epoch": 0.4616122457136006, "grad_norm": 0.16743382811546326, "learning_rate": 9.758497591738056e-05, "loss": 0.045, "step": 23100 }, { "epoch": 0.46181207785460215, "grad_norm": 0.12299620360136032, "learning_rate": 9.758180198559706e-05, "loss": 0.0434, "step": 23110 }, { "epoch": 0.4620119099956037, "grad_norm": 0.20570319890975952, "learning_rate": 9.757862602120638e-05, "loss": 0.0435, "step": 23120 }, { "epoch": 0.46221174213660526, "grad_norm": 0.16097593307495117, "learning_rate": 9.757544802434421e-05, "loss": 0.0492, "step": 23130 }, { "epoch": 0.4624115742776068, "grad_norm": 0.1471371352672577, "learning_rate": 9.757226799514629e-05, "loss": 0.0379, "step": 23140 }, { "epoch": 0.4626114064186084, "grad_norm": 0.11796779930591583, "learning_rate": 9.75690859337485e-05, "loss": 0.0375, "step": 23150 }, { "epoch": 0.4628112385596099, "grad_norm": 0.17434784770011902, "learning_rate": 9.756590184028671e-05, "loss": 0.0405, "step": 23160 }, { "epoch": 0.4630110707006115, "grad_norm": 0.15972983837127686, "learning_rate": 9.756271571489699e-05, "loss": 0.0485, "step": 23170 }, { "epoch": 0.46321090284161304, "grad_norm": 0.15021586418151855, "learning_rate": 9.755952755771544e-05, "loss": 0.0421, "step": 23180 }, { "epoch": 0.4634107349826146, "grad_norm": 0.14370714128017426, "learning_rate": 9.755633736887823e-05, "loss": 0.0411, "step": 23190 }, { "epoch": 0.46361056712361615, "grad_norm": 0.15449047088623047, "learning_rate": 9.755314514852165e-05, "loss": 0.0429, "step": 23200 }, { "epoch": 0.4638103992646177, "grad_norm": 0.10721572488546371, "learning_rate": 9.754995089678207e-05, "loss": 0.047, "step": 23210 }, { "epoch": 0.46401023140561926, "grad_norm": 0.18502439558506012, "learning_rate": 9.754675461379595e-05, "loss": 0.0395, "step": 23220 }, { "epoch": 0.46421006354662087, "grad_norm": 0.14728997647762299, "learning_rate": 9.754355629969981e-05, "loss": 0.0441, "step": 23230 }, { "epoch": 0.4644098956876224, "grad_norm": 0.16676868498325348, "learning_rate": 9.754035595463029e-05, "loss": 0.0507, "step": 23240 }, { "epoch": 0.464609727828624, "grad_norm": 0.1476396769285202, "learning_rate": 9.753715357872409e-05, "loss": 0.0388, "step": 23250 }, { "epoch": 0.46480955996962553, "grad_norm": 0.13306881487369537, "learning_rate": 9.753394917211802e-05, "loss": 0.0416, "step": 23260 }, { "epoch": 0.4650093921106271, "grad_norm": 0.13868829607963562, "learning_rate": 9.753074273494898e-05, "loss": 0.0519, "step": 23270 }, { "epoch": 0.46520922425162864, "grad_norm": 0.10629799216985703, "learning_rate": 9.75275342673539e-05, "loss": 0.0313, "step": 23280 }, { "epoch": 0.4654090563926302, "grad_norm": 0.14710412919521332, "learning_rate": 9.752432376946988e-05, "loss": 0.0454, "step": 23290 }, { "epoch": 0.46560888853363175, "grad_norm": 0.12541916966438293, "learning_rate": 9.752111124143406e-05, "loss": 0.0344, "step": 23300 }, { "epoch": 0.4658087206746333, "grad_norm": 0.12180943042039871, "learning_rate": 9.751789668338365e-05, "loss": 0.0457, "step": 23310 }, { "epoch": 0.46600855281563486, "grad_norm": 0.12922319769859314, "learning_rate": 9.751468009545601e-05, "loss": 0.0363, "step": 23320 }, { "epoch": 0.4662083849566364, "grad_norm": 0.1258969008922577, "learning_rate": 9.751146147778851e-05, "loss": 0.0452, "step": 23330 }, { "epoch": 0.46640821709763797, "grad_norm": 0.18407998979091644, "learning_rate": 9.750824083051866e-05, "loss": 0.0435, "step": 23340 }, { "epoch": 0.4666080492386395, "grad_norm": 0.17146125435829163, "learning_rate": 9.750501815378403e-05, "loss": 0.0412, "step": 23350 }, { "epoch": 0.4668078813796411, "grad_norm": 0.1515679508447647, "learning_rate": 9.75017934477223e-05, "loss": 0.0405, "step": 23360 }, { "epoch": 0.4670077135206427, "grad_norm": 0.11628050357103348, "learning_rate": 9.749856671247122e-05, "loss": 0.0394, "step": 23370 }, { "epoch": 0.46720754566164424, "grad_norm": 0.13272351026535034, "learning_rate": 9.749533794816862e-05, "loss": 0.0471, "step": 23380 }, { "epoch": 0.4674073778026458, "grad_norm": 0.1373467743396759, "learning_rate": 9.749210715495244e-05, "loss": 0.0479, "step": 23390 }, { "epoch": 0.46760720994364735, "grad_norm": 0.13909116387367249, "learning_rate": 9.748887433296067e-05, "loss": 0.0414, "step": 23400 }, { "epoch": 0.4678070420846489, "grad_norm": 0.1450243592262268, "learning_rate": 9.748563948233147e-05, "loss": 0.0551, "step": 23410 }, { "epoch": 0.46800687422565046, "grad_norm": 0.1223030686378479, "learning_rate": 9.748240260320296e-05, "loss": 0.0397, "step": 23420 }, { "epoch": 0.468206706366652, "grad_norm": 0.14816229045391083, "learning_rate": 9.747916369571344e-05, "loss": 0.0384, "step": 23430 }, { "epoch": 0.46840653850765357, "grad_norm": 0.1286984384059906, "learning_rate": 9.747592276000126e-05, "loss": 0.0397, "step": 23440 }, { "epoch": 0.4686063706486551, "grad_norm": 0.15596216917037964, "learning_rate": 9.747267979620489e-05, "loss": 0.0379, "step": 23450 }, { "epoch": 0.4688062027896567, "grad_norm": 0.17310577630996704, "learning_rate": 9.746943480446285e-05, "loss": 0.0459, "step": 23460 }, { "epoch": 0.46900603493065823, "grad_norm": 0.14736777544021606, "learning_rate": 9.746618778491374e-05, "loss": 0.0437, "step": 23470 }, { "epoch": 0.4692058670716598, "grad_norm": 0.1493529975414276, "learning_rate": 9.74629387376963e-05, "loss": 0.0379, "step": 23480 }, { "epoch": 0.46940569921266134, "grad_norm": 0.15728244185447693, "learning_rate": 9.74596876629493e-05, "loss": 0.0431, "step": 23490 }, { "epoch": 0.4696055313536629, "grad_norm": 0.1256096065044403, "learning_rate": 9.745643456081163e-05, "loss": 0.0434, "step": 23500 }, { "epoch": 0.4698053634946645, "grad_norm": 0.11002219468355179, "learning_rate": 9.745317943142227e-05, "loss": 0.0444, "step": 23510 }, { "epoch": 0.47000519563566606, "grad_norm": 0.17057108879089355, "learning_rate": 9.744992227492026e-05, "loss": 0.0429, "step": 23520 }, { "epoch": 0.4702050277766676, "grad_norm": 0.1572413146495819, "learning_rate": 9.744666309144472e-05, "loss": 0.0322, "step": 23530 }, { "epoch": 0.47040485991766917, "grad_norm": 0.12758156657218933, "learning_rate": 9.744340188113492e-05, "loss": 0.0403, "step": 23540 }, { "epoch": 0.4706046920586707, "grad_norm": 0.130445659160614, "learning_rate": 9.744013864413013e-05, "loss": 0.0421, "step": 23550 }, { "epoch": 0.4708045241996723, "grad_norm": 0.12862107157707214, "learning_rate": 9.743687338056978e-05, "loss": 0.0359, "step": 23560 }, { "epoch": 0.47100435634067384, "grad_norm": 0.12581460177898407, "learning_rate": 9.743360609059332e-05, "loss": 0.0451, "step": 23570 }, { "epoch": 0.4712041884816754, "grad_norm": 0.15619775652885437, "learning_rate": 9.743033677434037e-05, "loss": 0.0422, "step": 23580 }, { "epoch": 0.47140402062267694, "grad_norm": 0.1668219417333603, "learning_rate": 9.742706543195055e-05, "loss": 0.0495, "step": 23590 }, { "epoch": 0.4716038527636785, "grad_norm": 0.2124674916267395, "learning_rate": 9.742379206356363e-05, "loss": 0.0355, "step": 23600 }, { "epoch": 0.47180368490468005, "grad_norm": 0.18912093341350555, "learning_rate": 9.742051666931943e-05, "loss": 0.035, "step": 23610 }, { "epoch": 0.4720035170456816, "grad_norm": 0.16168907284736633, "learning_rate": 9.741723924935787e-05, "loss": 0.0477, "step": 23620 }, { "epoch": 0.47220334918668316, "grad_norm": 0.17970708012580872, "learning_rate": 9.741395980381895e-05, "loss": 0.0512, "step": 23630 }, { "epoch": 0.4724031813276847, "grad_norm": 0.1986837089061737, "learning_rate": 9.741067833284278e-05, "loss": 0.0403, "step": 23640 }, { "epoch": 0.47260301346868633, "grad_norm": 0.14833053946495056, "learning_rate": 9.740739483656954e-05, "loss": 0.0477, "step": 23650 }, { "epoch": 0.4728028456096879, "grad_norm": 0.1754767894744873, "learning_rate": 9.740410931513946e-05, "loss": 0.0435, "step": 23660 }, { "epoch": 0.47300267775068944, "grad_norm": 0.17021481692790985, "learning_rate": 9.740082176869292e-05, "loss": 0.0363, "step": 23670 }, { "epoch": 0.473202509891691, "grad_norm": 0.16451160609722137, "learning_rate": 9.739753219737036e-05, "loss": 0.0435, "step": 23680 }, { "epoch": 0.47340234203269255, "grad_norm": 0.13635268807411194, "learning_rate": 9.739424060131231e-05, "loss": 0.0456, "step": 23690 }, { "epoch": 0.4736021741736941, "grad_norm": 0.16148963570594788, "learning_rate": 9.739094698065934e-05, "loss": 0.0397, "step": 23700 }, { "epoch": 0.47380200631469566, "grad_norm": 0.1543160229921341, "learning_rate": 9.738765133555219e-05, "loss": 0.0477, "step": 23710 }, { "epoch": 0.4740018384556972, "grad_norm": 0.15406353771686554, "learning_rate": 9.738435366613162e-05, "loss": 0.0498, "step": 23720 }, { "epoch": 0.47420167059669877, "grad_norm": 0.1471903920173645, "learning_rate": 9.738105397253851e-05, "loss": 0.0468, "step": 23730 }, { "epoch": 0.4744015027377003, "grad_norm": 0.16781826317310333, "learning_rate": 9.737775225491382e-05, "loss": 0.0522, "step": 23740 }, { "epoch": 0.4746013348787019, "grad_norm": 0.15242744982242584, "learning_rate": 9.737444851339859e-05, "loss": 0.0391, "step": 23750 }, { "epoch": 0.47480116701970343, "grad_norm": 0.13696593046188354, "learning_rate": 9.737114274813395e-05, "loss": 0.0404, "step": 23760 }, { "epoch": 0.475000999160705, "grad_norm": 0.14995424449443817, "learning_rate": 9.736783495926112e-05, "loss": 0.0372, "step": 23770 }, { "epoch": 0.47520083130170654, "grad_norm": 0.15849599242210388, "learning_rate": 9.736452514692138e-05, "loss": 0.0533, "step": 23780 }, { "epoch": 0.47540066344270815, "grad_norm": 0.1966468095779419, "learning_rate": 9.736121331125616e-05, "loss": 0.0476, "step": 23790 }, { "epoch": 0.4756004955837097, "grad_norm": 0.151123508810997, "learning_rate": 9.735789945240689e-05, "loss": 0.0467, "step": 23800 }, { "epoch": 0.47580032772471126, "grad_norm": 0.15841753780841827, "learning_rate": 9.735458357051517e-05, "loss": 0.041, "step": 23810 }, { "epoch": 0.4760001598657128, "grad_norm": 0.14320194721221924, "learning_rate": 9.735126566572264e-05, "loss": 0.0446, "step": 23820 }, { "epoch": 0.47619999200671437, "grad_norm": 0.12837737798690796, "learning_rate": 9.7347945738171e-05, "loss": 0.0404, "step": 23830 }, { "epoch": 0.4763998241477159, "grad_norm": 0.1457175314426422, "learning_rate": 9.734462378800214e-05, "loss": 0.0423, "step": 23840 }, { "epoch": 0.4765996562887175, "grad_norm": 0.14190004765987396, "learning_rate": 9.734129981535789e-05, "loss": 0.038, "step": 23850 }, { "epoch": 0.47679948842971903, "grad_norm": 0.1433434933423996, "learning_rate": 9.733797382038028e-05, "loss": 0.0418, "step": 23860 }, { "epoch": 0.4769993205707206, "grad_norm": 0.15121857821941376, "learning_rate": 9.73346458032114e-05, "loss": 0.0475, "step": 23870 }, { "epoch": 0.47719915271172214, "grad_norm": 0.17513640224933624, "learning_rate": 9.733131576399342e-05, "loss": 0.0419, "step": 23880 }, { "epoch": 0.4773989848527237, "grad_norm": 0.1923220157623291, "learning_rate": 9.732798370286858e-05, "loss": 0.0504, "step": 23890 }, { "epoch": 0.47759881699372525, "grad_norm": 0.12452102452516556, "learning_rate": 9.732464961997921e-05, "loss": 0.0348, "step": 23900 }, { "epoch": 0.4777986491347268, "grad_norm": 0.16809538006782532, "learning_rate": 9.732131351546775e-05, "loss": 0.0388, "step": 23910 }, { "epoch": 0.4779984812757284, "grad_norm": 0.1344193071126938, "learning_rate": 9.731797538947671e-05, "loss": 0.0416, "step": 23920 }, { "epoch": 0.47819831341672997, "grad_norm": 0.19693416357040405, "learning_rate": 9.731463524214868e-05, "loss": 0.0516, "step": 23930 }, { "epoch": 0.4783981455577315, "grad_norm": 0.16032065451145172, "learning_rate": 9.731129307362637e-05, "loss": 0.0391, "step": 23940 }, { "epoch": 0.4785979776987331, "grad_norm": 0.17701736092567444, "learning_rate": 9.730794888405252e-05, "loss": 0.0425, "step": 23950 }, { "epoch": 0.47879780983973463, "grad_norm": 0.14012457430362701, "learning_rate": 9.730460267357e-05, "loss": 0.0424, "step": 23960 }, { "epoch": 0.4789976419807362, "grad_norm": 0.1333594024181366, "learning_rate": 9.730125444232175e-05, "loss": 0.0387, "step": 23970 }, { "epoch": 0.47919747412173774, "grad_norm": 0.10818289965391159, "learning_rate": 9.729790419045081e-05, "loss": 0.0397, "step": 23980 }, { "epoch": 0.4793973062627393, "grad_norm": 0.15649518370628357, "learning_rate": 9.729455191810031e-05, "loss": 0.0376, "step": 23990 }, { "epoch": 0.47959713840374085, "grad_norm": 0.14667929708957672, "learning_rate": 9.729119762541342e-05, "loss": 0.0415, "step": 24000 }, { "epoch": 0.4797969705447424, "grad_norm": 0.2155291736125946, "learning_rate": 9.728784131253344e-05, "loss": 0.046, "step": 24010 }, { "epoch": 0.47999680268574396, "grad_norm": 0.15730412304401398, "learning_rate": 9.728448297960375e-05, "loss": 0.0426, "step": 24020 }, { "epoch": 0.4801966348267455, "grad_norm": 0.1481262892484665, "learning_rate": 9.728112262676782e-05, "loss": 0.0406, "step": 24030 }, { "epoch": 0.48039646696774707, "grad_norm": 0.15283530950546265, "learning_rate": 9.727776025416919e-05, "loss": 0.0484, "step": 24040 }, { "epoch": 0.4805962991087486, "grad_norm": 0.16567376255989075, "learning_rate": 9.727439586195149e-05, "loss": 0.0433, "step": 24050 }, { "epoch": 0.48079613124975024, "grad_norm": 0.162244975566864, "learning_rate": 9.727102945025843e-05, "loss": 0.0451, "step": 24060 }, { "epoch": 0.4809959633907518, "grad_norm": 0.1496141105890274, "learning_rate": 9.726766101923385e-05, "loss": 0.0367, "step": 24070 }, { "epoch": 0.48119579553175335, "grad_norm": 0.19923555850982666, "learning_rate": 9.726429056902163e-05, "loss": 0.0456, "step": 24080 }, { "epoch": 0.4813956276727549, "grad_norm": 0.15823599696159363, "learning_rate": 9.726091809976571e-05, "loss": 0.0487, "step": 24090 }, { "epoch": 0.48159545981375645, "grad_norm": 0.15408855676651, "learning_rate": 9.725754361161023e-05, "loss": 0.0466, "step": 24100 }, { "epoch": 0.481795291954758, "grad_norm": 0.1308055967092514, "learning_rate": 9.725416710469927e-05, "loss": 0.035, "step": 24110 }, { "epoch": 0.48199512409575956, "grad_norm": 0.1408158242702484, "learning_rate": 9.725078857917713e-05, "loss": 0.0586, "step": 24120 }, { "epoch": 0.4821949562367611, "grad_norm": 0.12866951525211334, "learning_rate": 9.724740803518807e-05, "loss": 0.0473, "step": 24130 }, { "epoch": 0.4823947883777627, "grad_norm": 0.18313293159008026, "learning_rate": 9.724402547287657e-05, "loss": 0.0384, "step": 24140 }, { "epoch": 0.48259462051876423, "grad_norm": 0.16612336039543152, "learning_rate": 9.724064089238708e-05, "loss": 0.0388, "step": 24150 }, { "epoch": 0.4827944526597658, "grad_norm": 0.17320163547992706, "learning_rate": 9.723725429386418e-05, "loss": 0.043, "step": 24160 }, { "epoch": 0.48299428480076734, "grad_norm": 0.13166695833206177, "learning_rate": 9.723386567745257e-05, "loss": 0.0349, "step": 24170 }, { "epoch": 0.4831941169417689, "grad_norm": 0.11692565679550171, "learning_rate": 9.723047504329699e-05, "loss": 0.0384, "step": 24180 }, { "epoch": 0.48339394908277045, "grad_norm": 0.11836592853069305, "learning_rate": 9.722708239154227e-05, "loss": 0.0506, "step": 24190 }, { "epoch": 0.48359378122377206, "grad_norm": 0.10178694128990173, "learning_rate": 9.722368772233337e-05, "loss": 0.0427, "step": 24200 }, { "epoch": 0.4837936133647736, "grad_norm": 0.1411956250667572, "learning_rate": 9.722029103581527e-05, "loss": 0.0385, "step": 24210 }, { "epoch": 0.48399344550577517, "grad_norm": 0.14516319334506989, "learning_rate": 9.721689233213307e-05, "loss": 0.0415, "step": 24220 }, { "epoch": 0.4841932776467767, "grad_norm": 0.14696161448955536, "learning_rate": 9.721349161143199e-05, "loss": 0.0408, "step": 24230 }, { "epoch": 0.4843931097877783, "grad_norm": 0.14294712245464325, "learning_rate": 9.721008887385727e-05, "loss": 0.0487, "step": 24240 }, { "epoch": 0.48459294192877983, "grad_norm": 0.17963296175003052, "learning_rate": 9.720668411955429e-05, "loss": 0.0396, "step": 24250 }, { "epoch": 0.4847927740697814, "grad_norm": 0.15102285146713257, "learning_rate": 9.720327734866847e-05, "loss": 0.0402, "step": 24260 }, { "epoch": 0.48499260621078294, "grad_norm": 0.1623845249414444, "learning_rate": 9.719986856134538e-05, "loss": 0.0389, "step": 24270 }, { "epoch": 0.4851924383517845, "grad_norm": 0.14067821204662323, "learning_rate": 9.71964577577306e-05, "loss": 0.0453, "step": 24280 }, { "epoch": 0.48539227049278605, "grad_norm": 0.13845227658748627, "learning_rate": 9.719304493796985e-05, "loss": 0.0495, "step": 24290 }, { "epoch": 0.4855921026337876, "grad_norm": 0.1625431478023529, "learning_rate": 9.718963010220892e-05, "loss": 0.0398, "step": 24300 }, { "epoch": 0.48579193477478916, "grad_norm": 0.1365879774093628, "learning_rate": 9.718621325059367e-05, "loss": 0.0573, "step": 24310 }, { "epoch": 0.4859917669157907, "grad_norm": 0.13519570231437683, "learning_rate": 9.71827943832701e-05, "loss": 0.0406, "step": 24320 }, { "epoch": 0.48619159905679227, "grad_norm": 0.1533336341381073, "learning_rate": 9.71793735003842e-05, "loss": 0.0417, "step": 24330 }, { "epoch": 0.4863914311977939, "grad_norm": 0.12296038866043091, "learning_rate": 9.717595060208216e-05, "loss": 0.0433, "step": 24340 }, { "epoch": 0.48659126333879543, "grad_norm": 0.15185251832008362, "learning_rate": 9.717252568851015e-05, "loss": 0.0487, "step": 24350 }, { "epoch": 0.486791095479797, "grad_norm": 0.13153304159641266, "learning_rate": 9.716909875981453e-05, "loss": 0.0426, "step": 24360 }, { "epoch": 0.48699092762079854, "grad_norm": 0.1416836529970169, "learning_rate": 9.716566981614163e-05, "loss": 0.0521, "step": 24370 }, { "epoch": 0.4871907597618001, "grad_norm": 0.16791847348213196, "learning_rate": 9.716223885763799e-05, "loss": 0.0426, "step": 24380 }, { "epoch": 0.48739059190280165, "grad_norm": 0.13593924045562744, "learning_rate": 9.715880588445014e-05, "loss": 0.0334, "step": 24390 }, { "epoch": 0.4875904240438032, "grad_norm": 0.13220474123954773, "learning_rate": 9.715537089672472e-05, "loss": 0.0411, "step": 24400 }, { "epoch": 0.48779025618480476, "grad_norm": 0.12786492705345154, "learning_rate": 9.71519338946085e-05, "loss": 0.0453, "step": 24410 }, { "epoch": 0.4879900883258063, "grad_norm": 0.13142260909080505, "learning_rate": 9.714849487824828e-05, "loss": 0.0529, "step": 24420 }, { "epoch": 0.48818992046680787, "grad_norm": 0.22107836604118347, "learning_rate": 9.714505384779098e-05, "loss": 0.044, "step": 24430 }, { "epoch": 0.4883897526078094, "grad_norm": 0.14668388664722443, "learning_rate": 9.714161080338358e-05, "loss": 0.0466, "step": 24440 }, { "epoch": 0.488589584748811, "grad_norm": 0.19519765675067902, "learning_rate": 9.713816574517316e-05, "loss": 0.0477, "step": 24450 }, { "epoch": 0.48878941688981253, "grad_norm": 0.1628526896238327, "learning_rate": 9.713471867330691e-05, "loss": 0.0512, "step": 24460 }, { "epoch": 0.48898924903081414, "grad_norm": 0.15804079174995422, "learning_rate": 9.713126958793206e-05, "loss": 0.0385, "step": 24470 }, { "epoch": 0.4891890811718157, "grad_norm": 0.18433700501918793, "learning_rate": 9.712781848919597e-05, "loss": 0.0455, "step": 24480 }, { "epoch": 0.48938891331281725, "grad_norm": 0.1443975567817688, "learning_rate": 9.712436537724602e-05, "loss": 0.0425, "step": 24490 }, { "epoch": 0.4895887454538188, "grad_norm": 0.15381012856960297, "learning_rate": 9.712091025222978e-05, "loss": 0.0395, "step": 24500 }, { "epoch": 0.48978857759482036, "grad_norm": 0.12156996130943298, "learning_rate": 9.711745311429481e-05, "loss": 0.0472, "step": 24510 }, { "epoch": 0.4899884097358219, "grad_norm": 0.14943931996822357, "learning_rate": 9.711399396358881e-05, "loss": 0.044, "step": 24520 }, { "epoch": 0.49018824187682347, "grad_norm": 0.13879002630710602, "learning_rate": 9.711053280025952e-05, "loss": 0.0445, "step": 24530 }, { "epoch": 0.490388074017825, "grad_norm": 0.14159680902957916, "learning_rate": 9.710706962445482e-05, "loss": 0.049, "step": 24540 }, { "epoch": 0.4905879061588266, "grad_norm": 0.14035561680793762, "learning_rate": 9.710360443632264e-05, "loss": 0.0405, "step": 24550 }, { "epoch": 0.49078773829982814, "grad_norm": 0.1765023171901703, "learning_rate": 9.710013723601101e-05, "loss": 0.048, "step": 24560 }, { "epoch": 0.4909875704408297, "grad_norm": 0.1637493073940277, "learning_rate": 9.709666802366806e-05, "loss": 0.0425, "step": 24570 }, { "epoch": 0.49118740258183125, "grad_norm": 0.14552830159664154, "learning_rate": 9.709319679944195e-05, "loss": 0.0403, "step": 24580 }, { "epoch": 0.4913872347228328, "grad_norm": 0.15569071471691132, "learning_rate": 9.7089723563481e-05, "loss": 0.0405, "step": 24590 }, { "epoch": 0.49158706686383435, "grad_norm": 0.20847581326961517, "learning_rate": 9.708624831593356e-05, "loss": 0.0409, "step": 24600 }, { "epoch": 0.49178689900483596, "grad_norm": 0.13002941012382507, "learning_rate": 9.70827710569481e-05, "loss": 0.0421, "step": 24610 }, { "epoch": 0.4919867311458375, "grad_norm": 0.11693049967288971, "learning_rate": 9.707929178667315e-05, "loss": 0.0394, "step": 24620 }, { "epoch": 0.4921865632868391, "grad_norm": 0.14912328124046326, "learning_rate": 9.707581050525734e-05, "loss": 0.0418, "step": 24630 }, { "epoch": 0.49238639542784063, "grad_norm": 0.181101992726326, "learning_rate": 9.707232721284938e-05, "loss": 0.041, "step": 24640 }, { "epoch": 0.4925862275688422, "grad_norm": 0.1338472068309784, "learning_rate": 9.706884190959808e-05, "loss": 0.0369, "step": 24650 }, { "epoch": 0.49278605970984374, "grad_norm": 0.12234822660684586, "learning_rate": 9.706535459565233e-05, "loss": 0.0423, "step": 24660 }, { "epoch": 0.4929858918508453, "grad_norm": 0.1286807805299759, "learning_rate": 9.706186527116107e-05, "loss": 0.0429, "step": 24670 }, { "epoch": 0.49318572399184685, "grad_norm": 0.11386866867542267, "learning_rate": 9.705837393627342e-05, "loss": 0.0378, "step": 24680 }, { "epoch": 0.4933855561328484, "grad_norm": 0.16894732415676117, "learning_rate": 9.705488059113846e-05, "loss": 0.0459, "step": 24690 }, { "epoch": 0.49358538827384996, "grad_norm": 0.17462904751300812, "learning_rate": 9.705138523590545e-05, "loss": 0.0406, "step": 24700 }, { "epoch": 0.4937852204148515, "grad_norm": 0.15835128724575043, "learning_rate": 9.70478878707237e-05, "loss": 0.0369, "step": 24710 }, { "epoch": 0.49398505255585307, "grad_norm": 0.12217502295970917, "learning_rate": 9.70443884957426e-05, "loss": 0.0408, "step": 24720 }, { "epoch": 0.4941848846968546, "grad_norm": 0.1302356719970703, "learning_rate": 9.704088711111165e-05, "loss": 0.036, "step": 24730 }, { "epoch": 0.4943847168378562, "grad_norm": 0.15329872071743011, "learning_rate": 9.703738371698042e-05, "loss": 0.035, "step": 24740 }, { "epoch": 0.4945845489788578, "grad_norm": 0.13781210780143738, "learning_rate": 9.703387831349857e-05, "loss": 0.0456, "step": 24750 }, { "epoch": 0.49478438111985934, "grad_norm": 0.14979764819145203, "learning_rate": 9.703037090081584e-05, "loss": 0.0512, "step": 24760 }, { "epoch": 0.4949842132608609, "grad_norm": 0.1202988252043724, "learning_rate": 9.702686147908207e-05, "loss": 0.0372, "step": 24770 }, { "epoch": 0.49518404540186245, "grad_norm": 0.1285301148891449, "learning_rate": 9.702335004844716e-05, "loss": 0.0438, "step": 24780 }, { "epoch": 0.495383877542864, "grad_norm": 0.17618124186992645, "learning_rate": 9.701983660906114e-05, "loss": 0.0477, "step": 24790 }, { "epoch": 0.49558370968386556, "grad_norm": 0.18849505484104156, "learning_rate": 9.701632116107405e-05, "loss": 0.0505, "step": 24800 }, { "epoch": 0.4957835418248671, "grad_norm": 0.1574689745903015, "learning_rate": 9.70128037046361e-05, "loss": 0.0473, "step": 24810 }, { "epoch": 0.49598337396586867, "grad_norm": 0.17371812462806702, "learning_rate": 9.700928423989755e-05, "loss": 0.0454, "step": 24820 }, { "epoch": 0.4961832061068702, "grad_norm": 0.1349545270204544, "learning_rate": 9.700576276700873e-05, "loss": 0.0439, "step": 24830 }, { "epoch": 0.4963830382478718, "grad_norm": 0.1290631741285324, "learning_rate": 9.700223928612008e-05, "loss": 0.0492, "step": 24840 }, { "epoch": 0.49658287038887333, "grad_norm": 0.09920378029346466, "learning_rate": 9.699871379738212e-05, "loss": 0.0464, "step": 24850 }, { "epoch": 0.4967827025298749, "grad_norm": 0.13040180504322052, "learning_rate": 9.699518630094544e-05, "loss": 0.0436, "step": 24860 }, { "epoch": 0.49698253467087644, "grad_norm": 0.14468082785606384, "learning_rate": 9.699165679696072e-05, "loss": 0.0388, "step": 24870 }, { "epoch": 0.497182366811878, "grad_norm": 0.11567641794681549, "learning_rate": 9.698812528557877e-05, "loss": 0.045, "step": 24880 }, { "epoch": 0.4973821989528796, "grad_norm": 0.13225269317626953, "learning_rate": 9.698459176695042e-05, "loss": 0.0351, "step": 24890 }, { "epoch": 0.49758203109388116, "grad_norm": 0.43881848454475403, "learning_rate": 9.698105624122661e-05, "loss": 0.04, "step": 24900 }, { "epoch": 0.4977818632348827, "grad_norm": 0.1277054250240326, "learning_rate": 9.69775187085584e-05, "loss": 0.0442, "step": 24910 }, { "epoch": 0.49798169537588427, "grad_norm": 0.20528624951839447, "learning_rate": 9.69739791690969e-05, "loss": 0.0471, "step": 24920 }, { "epoch": 0.4981815275168858, "grad_norm": 0.1640242338180542, "learning_rate": 9.697043762299328e-05, "loss": 0.0451, "step": 24930 }, { "epoch": 0.4983813596578874, "grad_norm": 0.15968453884124756, "learning_rate": 9.696689407039886e-05, "loss": 0.039, "step": 24940 }, { "epoch": 0.49858119179888893, "grad_norm": 0.15687888860702515, "learning_rate": 9.696334851146501e-05, "loss": 0.0506, "step": 24950 }, { "epoch": 0.4987810239398905, "grad_norm": 0.12871597707271576, "learning_rate": 9.695980094634318e-05, "loss": 0.0417, "step": 24960 }, { "epoch": 0.49898085608089204, "grad_norm": 0.13239562511444092, "learning_rate": 9.695625137518493e-05, "loss": 0.0389, "step": 24970 }, { "epoch": 0.4991806882218936, "grad_norm": 0.13574354350566864, "learning_rate": 9.695269979814188e-05, "loss": 0.0461, "step": 24980 }, { "epoch": 0.49938052036289515, "grad_norm": 0.14411862194538116, "learning_rate": 9.694914621536575e-05, "loss": 0.0377, "step": 24990 }, { "epoch": 0.4995803525038967, "grad_norm": 0.17000029981136322, "learning_rate": 9.694559062700835e-05, "loss": 0.0449, "step": 25000 }, { "epoch": 0.49978018464489826, "grad_norm": 0.15570470690727234, "learning_rate": 9.694203303322155e-05, "loss": 0.0367, "step": 25010 }, { "epoch": 0.4999800167858998, "grad_norm": 0.15427939593791962, "learning_rate": 9.693847343415734e-05, "loss": 0.0421, "step": 25020 }, { "epoch": 0.5001798489269014, "grad_norm": 0.16730882227420807, "learning_rate": 9.693491182996776e-05, "loss": 0.044, "step": 25030 }, { "epoch": 0.500379681067903, "grad_norm": 0.14142931997776031, "learning_rate": 9.693134822080499e-05, "loss": 0.0385, "step": 25040 }, { "epoch": 0.5005795132089045, "grad_norm": 0.12779028713703156, "learning_rate": 9.692778260682121e-05, "loss": 0.04, "step": 25050 }, { "epoch": 0.5007793453499061, "grad_norm": 0.21098576486110687, "learning_rate": 9.69242149881688e-05, "loss": 0.0395, "step": 25060 }, { "epoch": 0.5009791774909076, "grad_norm": 0.13653269410133362, "learning_rate": 9.69206453650001e-05, "loss": 0.0409, "step": 25070 }, { "epoch": 0.5011790096319092, "grad_norm": 0.13505682349205017, "learning_rate": 9.691707373746764e-05, "loss": 0.0424, "step": 25080 }, { "epoch": 0.5013788417729108, "grad_norm": 0.17592468857765198, "learning_rate": 9.691350010572398e-05, "loss": 0.0386, "step": 25090 }, { "epoch": 0.5015786739139123, "grad_norm": 0.1140260398387909, "learning_rate": 9.690992446992177e-05, "loss": 0.0442, "step": 25100 }, { "epoch": 0.5017785060549139, "grad_norm": 0.12946456670761108, "learning_rate": 9.690634683021377e-05, "loss": 0.0383, "step": 25110 }, { "epoch": 0.5019783381959154, "grad_norm": 0.16042965650558472, "learning_rate": 9.69027671867528e-05, "loss": 0.036, "step": 25120 }, { "epoch": 0.502178170336917, "grad_norm": 0.1470899134874344, "learning_rate": 9.689918553969179e-05, "loss": 0.0399, "step": 25130 }, { "epoch": 0.5023780024779185, "grad_norm": 0.13956670463085175, "learning_rate": 9.689560188918372e-05, "loss": 0.0511, "step": 25140 }, { "epoch": 0.5025778346189201, "grad_norm": 0.16068369150161743, "learning_rate": 9.689201623538168e-05, "loss": 0.0427, "step": 25150 }, { "epoch": 0.5027776667599216, "grad_norm": 0.13391923904418945, "learning_rate": 9.688842857843886e-05, "loss": 0.0442, "step": 25160 }, { "epoch": 0.5029774989009232, "grad_norm": 0.17150169610977173, "learning_rate": 9.68848389185085e-05, "loss": 0.0442, "step": 25170 }, { "epoch": 0.5031773310419247, "grad_norm": 0.14214950799942017, "learning_rate": 9.688124725574397e-05, "loss": 0.0367, "step": 25180 }, { "epoch": 0.5033771631829264, "grad_norm": 0.10785799473524094, "learning_rate": 9.687765359029865e-05, "loss": 0.0419, "step": 25190 }, { "epoch": 0.5035769953239279, "grad_norm": 0.13523335754871368, "learning_rate": 9.687405792232611e-05, "loss": 0.0473, "step": 25200 }, { "epoch": 0.5037768274649295, "grad_norm": 0.1562950313091278, "learning_rate": 9.687046025197992e-05, "loss": 0.0378, "step": 25210 }, { "epoch": 0.503976659605931, "grad_norm": 0.13821882009506226, "learning_rate": 9.686686057941378e-05, "loss": 0.0386, "step": 25220 }, { "epoch": 0.5041764917469326, "grad_norm": 0.15448828041553497, "learning_rate": 9.686325890478144e-05, "loss": 0.0475, "step": 25230 }, { "epoch": 0.5043763238879342, "grad_norm": 0.12252704054117203, "learning_rate": 9.685965522823677e-05, "loss": 0.0371, "step": 25240 }, { "epoch": 0.5045761560289357, "grad_norm": 0.10664772987365723, "learning_rate": 9.685604954993373e-05, "loss": 0.0381, "step": 25250 }, { "epoch": 0.5047759881699373, "grad_norm": 0.12346798926591873, "learning_rate": 9.685244187002631e-05, "loss": 0.0465, "step": 25260 }, { "epoch": 0.5049758203109388, "grad_norm": 0.16195259988307953, "learning_rate": 9.684883218866865e-05, "loss": 0.0429, "step": 25270 }, { "epoch": 0.5051756524519404, "grad_norm": 0.13105064630508423, "learning_rate": 9.684522050601495e-05, "loss": 0.0397, "step": 25280 }, { "epoch": 0.5053754845929419, "grad_norm": 0.1191854327917099, "learning_rate": 9.684160682221947e-05, "loss": 0.0424, "step": 25290 }, { "epoch": 0.5055753167339435, "grad_norm": 0.12663985788822174, "learning_rate": 9.683799113743662e-05, "loss": 0.0417, "step": 25300 }, { "epoch": 0.505775148874945, "grad_norm": 0.1587827503681183, "learning_rate": 9.683437345182082e-05, "loss": 0.0418, "step": 25310 }, { "epoch": 0.5059749810159466, "grad_norm": 0.16449283063411713, "learning_rate": 9.683075376552663e-05, "loss": 0.0467, "step": 25320 }, { "epoch": 0.5061748131569481, "grad_norm": 0.14682798087596893, "learning_rate": 9.682713207870866e-05, "loss": 0.0394, "step": 25330 }, { "epoch": 0.5063746452979497, "grad_norm": 0.15337203443050385, "learning_rate": 9.682350839152165e-05, "loss": 0.0517, "step": 25340 }, { "epoch": 0.5065744774389512, "grad_norm": 0.15672878921031952, "learning_rate": 9.681988270412035e-05, "loss": 0.0395, "step": 25350 }, { "epoch": 0.5067743095799528, "grad_norm": 0.1574544608592987, "learning_rate": 9.68162550166597e-05, "loss": 0.0356, "step": 25360 }, { "epoch": 0.5069741417209545, "grad_norm": 0.13738563656806946, "learning_rate": 9.681262532929462e-05, "loss": 0.0501, "step": 25370 }, { "epoch": 0.507173973861956, "grad_norm": 0.15917575359344482, "learning_rate": 9.68089936421802e-05, "loss": 0.0514, "step": 25380 }, { "epoch": 0.5073738060029576, "grad_norm": 0.1652631163597107, "learning_rate": 9.680535995547155e-05, "loss": 0.0474, "step": 25390 }, { "epoch": 0.5075736381439591, "grad_norm": 0.13259071111679077, "learning_rate": 9.680172426932391e-05, "loss": 0.0451, "step": 25400 }, { "epoch": 0.5077734702849607, "grad_norm": 0.1605498492717743, "learning_rate": 9.679808658389259e-05, "loss": 0.0367, "step": 25410 }, { "epoch": 0.5079733024259622, "grad_norm": 0.13471092283725739, "learning_rate": 9.679444689933299e-05, "loss": 0.0365, "step": 25420 }, { "epoch": 0.5081731345669638, "grad_norm": 0.14036965370178223, "learning_rate": 9.679080521580057e-05, "loss": 0.0448, "step": 25430 }, { "epoch": 0.5083729667079653, "grad_norm": 0.18215572834014893, "learning_rate": 9.678716153345092e-05, "loss": 0.0463, "step": 25440 }, { "epoch": 0.5085727988489669, "grad_norm": 0.14462077617645264, "learning_rate": 9.678351585243966e-05, "loss": 0.0414, "step": 25450 }, { "epoch": 0.5087726309899684, "grad_norm": 0.20404739677906036, "learning_rate": 9.677986817292255e-05, "loss": 0.0451, "step": 25460 }, { "epoch": 0.50897246313097, "grad_norm": 0.13577623665332794, "learning_rate": 9.677621849505543e-05, "loss": 0.0373, "step": 25470 }, { "epoch": 0.5091722952719715, "grad_norm": 0.15063583850860596, "learning_rate": 9.677256681899417e-05, "loss": 0.0485, "step": 25480 }, { "epoch": 0.5093721274129731, "grad_norm": 0.15084081888198853, "learning_rate": 9.676891314489477e-05, "loss": 0.0401, "step": 25490 }, { "epoch": 0.5095719595539746, "grad_norm": 0.14320692420005798, "learning_rate": 9.676525747291334e-05, "loss": 0.0407, "step": 25500 }, { "epoch": 0.5097717916949762, "grad_norm": 0.18414540588855743, "learning_rate": 9.676159980320602e-05, "loss": 0.0454, "step": 25510 }, { "epoch": 0.5099716238359778, "grad_norm": 0.15068891644477844, "learning_rate": 9.675794013592905e-05, "loss": 0.0419, "step": 25520 }, { "epoch": 0.5101714559769793, "grad_norm": 0.14163067936897278, "learning_rate": 9.675427847123877e-05, "loss": 0.0476, "step": 25530 }, { "epoch": 0.5103712881179809, "grad_norm": 0.14008913934230804, "learning_rate": 9.675061480929162e-05, "loss": 0.0315, "step": 25540 }, { "epoch": 0.5105711202589824, "grad_norm": 0.15227529406547546, "learning_rate": 9.674694915024408e-05, "loss": 0.0527, "step": 25550 }, { "epoch": 0.510770952399984, "grad_norm": 0.15855231881141663, "learning_rate": 9.674328149425274e-05, "loss": 0.0401, "step": 25560 }, { "epoch": 0.5109707845409855, "grad_norm": 0.19277413189411163, "learning_rate": 9.673961184147429e-05, "loss": 0.0429, "step": 25570 }, { "epoch": 0.5111706166819872, "grad_norm": 0.13552679121494293, "learning_rate": 9.67359401920655e-05, "loss": 0.0482, "step": 25580 }, { "epoch": 0.5113704488229887, "grad_norm": 0.1528494656085968, "learning_rate": 9.673226654618319e-05, "loss": 0.0458, "step": 25590 }, { "epoch": 0.5115702809639903, "grad_norm": 0.19242818653583527, "learning_rate": 9.672859090398431e-05, "loss": 0.0413, "step": 25600 }, { "epoch": 0.5117701131049918, "grad_norm": 0.11610376089811325, "learning_rate": 9.672491326562586e-05, "loss": 0.0522, "step": 25610 }, { "epoch": 0.5119699452459934, "grad_norm": 0.13293994963169098, "learning_rate": 9.672123363126495e-05, "loss": 0.0454, "step": 25620 }, { "epoch": 0.5121697773869949, "grad_norm": 0.12440695613622665, "learning_rate": 9.671755200105877e-05, "loss": 0.042, "step": 25630 }, { "epoch": 0.5123696095279965, "grad_norm": 0.18287035822868347, "learning_rate": 9.67138683751646e-05, "loss": 0.0474, "step": 25640 }, { "epoch": 0.5125694416689981, "grad_norm": 0.20414702594280243, "learning_rate": 9.671018275373978e-05, "loss": 0.0414, "step": 25650 }, { "epoch": 0.5127692738099996, "grad_norm": 0.1725197285413742, "learning_rate": 9.670649513694177e-05, "loss": 0.048, "step": 25660 }, { "epoch": 0.5129691059510012, "grad_norm": 0.14876611530780792, "learning_rate": 9.67028055249281e-05, "loss": 0.0472, "step": 25670 }, { "epoch": 0.5131689380920027, "grad_norm": 0.1621427834033966, "learning_rate": 9.669911391785635e-05, "loss": 0.0444, "step": 25680 }, { "epoch": 0.5133687702330043, "grad_norm": 0.12398163974285126, "learning_rate": 9.669542031588425e-05, "loss": 0.0359, "step": 25690 }, { "epoch": 0.5135686023740058, "grad_norm": 0.14429861307144165, "learning_rate": 9.669172471916959e-05, "loss": 0.0423, "step": 25700 }, { "epoch": 0.5137684345150074, "grad_norm": 0.1644875705242157, "learning_rate": 9.668802712787022e-05, "loss": 0.0386, "step": 25710 }, { "epoch": 0.5139682666560089, "grad_norm": 0.1655529886484146, "learning_rate": 9.66843275421441e-05, "loss": 0.0477, "step": 25720 }, { "epoch": 0.5141680987970105, "grad_norm": 0.12942640483379364, "learning_rate": 9.668062596214925e-05, "loss": 0.0402, "step": 25730 }, { "epoch": 0.514367930938012, "grad_norm": 0.14871346950531006, "learning_rate": 9.667692238804383e-05, "loss": 0.0497, "step": 25740 }, { "epoch": 0.5145677630790136, "grad_norm": 0.12788279354572296, "learning_rate": 9.667321681998604e-05, "loss": 0.0413, "step": 25750 }, { "epoch": 0.5147675952200151, "grad_norm": 0.15470001101493835, "learning_rate": 9.666950925813416e-05, "loss": 0.0426, "step": 25760 }, { "epoch": 0.5149674273610167, "grad_norm": 0.15158064663410187, "learning_rate": 9.666579970264658e-05, "loss": 0.0406, "step": 25770 }, { "epoch": 0.5151672595020184, "grad_norm": 0.14350953698158264, "learning_rate": 9.666208815368176e-05, "loss": 0.045, "step": 25780 }, { "epoch": 0.5153670916430199, "grad_norm": 0.20036542415618896, "learning_rate": 9.665837461139827e-05, "loss": 0.0444, "step": 25790 }, { "epoch": 0.5155669237840215, "grad_norm": 0.11737390607595444, "learning_rate": 9.665465907595471e-05, "loss": 0.0342, "step": 25800 }, { "epoch": 0.515766755925023, "grad_norm": 0.1789722442626953, "learning_rate": 9.665094154750983e-05, "loss": 0.0435, "step": 25810 }, { "epoch": 0.5159665880660246, "grad_norm": 0.1549926996231079, "learning_rate": 9.664722202622242e-05, "loss": 0.0382, "step": 25820 }, { "epoch": 0.5161664202070261, "grad_norm": 0.14914783835411072, "learning_rate": 9.66435005122514e-05, "loss": 0.0491, "step": 25830 }, { "epoch": 0.5163662523480277, "grad_norm": 0.15051570534706116, "learning_rate": 9.663977700575571e-05, "loss": 0.0472, "step": 25840 }, { "epoch": 0.5165660844890292, "grad_norm": 0.13242276012897491, "learning_rate": 9.663605150689443e-05, "loss": 0.0407, "step": 25850 }, { "epoch": 0.5167659166300308, "grad_norm": 0.12637962400913239, "learning_rate": 9.663232401582671e-05, "loss": 0.0425, "step": 25860 }, { "epoch": 0.5169657487710323, "grad_norm": 0.15566739439964294, "learning_rate": 9.662859453271176e-05, "loss": 0.0349, "step": 25870 }, { "epoch": 0.5171655809120339, "grad_norm": 0.19039776921272278, "learning_rate": 9.662486305770892e-05, "loss": 0.0372, "step": 25880 }, { "epoch": 0.5173654130530354, "grad_norm": 0.13713541626930237, "learning_rate": 9.662112959097758e-05, "loss": 0.0436, "step": 25890 }, { "epoch": 0.517565245194037, "grad_norm": 0.15273460745811462, "learning_rate": 9.661739413267723e-05, "loss": 0.0416, "step": 25900 }, { "epoch": 0.5177650773350385, "grad_norm": 0.15998993813991547, "learning_rate": 9.661365668296745e-05, "loss": 0.0422, "step": 25910 }, { "epoch": 0.5179649094760401, "grad_norm": 0.11017061024904251, "learning_rate": 9.660991724200788e-05, "loss": 0.0411, "step": 25920 }, { "epoch": 0.5181647416170417, "grad_norm": 0.12949971854686737, "learning_rate": 9.660617580995828e-05, "loss": 0.0397, "step": 25930 }, { "epoch": 0.5183645737580432, "grad_norm": 0.1544971466064453, "learning_rate": 9.660243238697844e-05, "loss": 0.0496, "step": 25940 }, { "epoch": 0.5185644058990448, "grad_norm": 0.17675462365150452, "learning_rate": 9.659868697322833e-05, "loss": 0.04, "step": 25950 }, { "epoch": 0.5187642380400463, "grad_norm": 0.127947136759758, "learning_rate": 9.65949395688679e-05, "loss": 0.0439, "step": 25960 }, { "epoch": 0.518964070181048, "grad_norm": 0.13437047600746155, "learning_rate": 9.659119017405726e-05, "loss": 0.0426, "step": 25970 }, { "epoch": 0.5191639023220495, "grad_norm": 0.15968993306159973, "learning_rate": 9.658743878895655e-05, "loss": 0.0448, "step": 25980 }, { "epoch": 0.5193637344630511, "grad_norm": 0.13269248604774475, "learning_rate": 9.658368541372607e-05, "loss": 0.0359, "step": 25990 }, { "epoch": 0.5195635666040526, "grad_norm": 0.1528652310371399, "learning_rate": 9.65799300485261e-05, "loss": 0.0483, "step": 26000 }, { "epoch": 0.5197633987450542, "grad_norm": 0.14367708563804626, "learning_rate": 9.657617269351708e-05, "loss": 0.0471, "step": 26010 }, { "epoch": 0.5199632308860557, "grad_norm": 0.1346343755722046, "learning_rate": 9.657241334885956e-05, "loss": 0.0436, "step": 26020 }, { "epoch": 0.5201630630270573, "grad_norm": 0.15026606619358063, "learning_rate": 9.656865201471407e-05, "loss": 0.0419, "step": 26030 }, { "epoch": 0.5203628951680588, "grad_norm": 0.12204719334840775, "learning_rate": 9.656488869124132e-05, "loss": 0.0444, "step": 26040 }, { "epoch": 0.5205627273090604, "grad_norm": 0.13771818578243256, "learning_rate": 9.656112337860208e-05, "loss": 0.0479, "step": 26050 }, { "epoch": 0.520762559450062, "grad_norm": 0.12012434005737305, "learning_rate": 9.655735607695718e-05, "loss": 0.0401, "step": 26060 }, { "epoch": 0.5209623915910635, "grad_norm": 0.11551178991794586, "learning_rate": 9.655358678646754e-05, "loss": 0.0425, "step": 26070 }, { "epoch": 0.5211622237320651, "grad_norm": 0.1382240206003189, "learning_rate": 9.654981550729421e-05, "loss": 0.0319, "step": 26080 }, { "epoch": 0.5213620558730666, "grad_norm": 0.18427596986293793, "learning_rate": 9.654604223959826e-05, "loss": 0.0411, "step": 26090 }, { "epoch": 0.5215618880140682, "grad_norm": 0.1288221776485443, "learning_rate": 9.654226698354092e-05, "loss": 0.0425, "step": 26100 }, { "epoch": 0.5217617201550697, "grad_norm": 0.1763158142566681, "learning_rate": 9.653848973928342e-05, "loss": 0.0426, "step": 26110 }, { "epoch": 0.5219615522960713, "grad_norm": 0.21061548590660095, "learning_rate": 9.653471050698714e-05, "loss": 0.0418, "step": 26120 }, { "epoch": 0.5221613844370728, "grad_norm": 0.16230793297290802, "learning_rate": 9.653092928681351e-05, "loss": 0.03, "step": 26130 }, { "epoch": 0.5223612165780744, "grad_norm": 0.17981396615505219, "learning_rate": 9.652714607892406e-05, "loss": 0.0345, "step": 26140 }, { "epoch": 0.5225610487190759, "grad_norm": 0.16533733904361725, "learning_rate": 9.652336088348039e-05, "loss": 0.0405, "step": 26150 }, { "epoch": 0.5227608808600775, "grad_norm": 0.15499721467494965, "learning_rate": 9.651957370064423e-05, "loss": 0.0435, "step": 26160 }, { "epoch": 0.522960713001079, "grad_norm": 0.13725604116916656, "learning_rate": 9.651578453057732e-05, "loss": 0.0497, "step": 26170 }, { "epoch": 0.5231605451420807, "grad_norm": 0.16014517843723297, "learning_rate": 9.651199337344157e-05, "loss": 0.0347, "step": 26180 }, { "epoch": 0.5233603772830822, "grad_norm": 0.19730739295482635, "learning_rate": 9.65082002293989e-05, "loss": 0.0468, "step": 26190 }, { "epoch": 0.5235602094240838, "grad_norm": 0.13965585827827454, "learning_rate": 9.650440509861135e-05, "loss": 0.0433, "step": 26200 }, { "epoch": 0.5237600415650854, "grad_norm": 0.14598187804222107, "learning_rate": 9.650060798124105e-05, "loss": 0.0446, "step": 26210 }, { "epoch": 0.5239598737060869, "grad_norm": 0.14152787625789642, "learning_rate": 9.64968088774502e-05, "loss": 0.0361, "step": 26220 }, { "epoch": 0.5241597058470885, "grad_norm": 0.2019331455230713, "learning_rate": 9.64930077874011e-05, "loss": 0.0397, "step": 26230 }, { "epoch": 0.52435953798809, "grad_norm": 0.15292610228061676, "learning_rate": 9.648920471125609e-05, "loss": 0.0464, "step": 26240 }, { "epoch": 0.5245593701290916, "grad_norm": 0.149289071559906, "learning_rate": 9.648539964917769e-05, "loss": 0.0472, "step": 26250 }, { "epoch": 0.5247592022700931, "grad_norm": 0.12685935199260712, "learning_rate": 9.648159260132839e-05, "loss": 0.0412, "step": 26260 }, { "epoch": 0.5249590344110947, "grad_norm": 0.13030022382736206, "learning_rate": 9.647778356787084e-05, "loss": 0.0405, "step": 26270 }, { "epoch": 0.5251588665520962, "grad_norm": 0.14207109808921814, "learning_rate": 9.647397254896779e-05, "loss": 0.0404, "step": 26280 }, { "epoch": 0.5253586986930978, "grad_norm": 0.11965026706457138, "learning_rate": 9.647015954478198e-05, "loss": 0.0421, "step": 26290 }, { "epoch": 0.5255585308340993, "grad_norm": 0.1262911707162857, "learning_rate": 9.646634455547633e-05, "loss": 0.0376, "step": 26300 }, { "epoch": 0.5257583629751009, "grad_norm": 0.1396748572587967, "learning_rate": 9.646252758121381e-05, "loss": 0.0339, "step": 26310 }, { "epoch": 0.5259581951161024, "grad_norm": 0.1276456117630005, "learning_rate": 9.645870862215747e-05, "loss": 0.0349, "step": 26320 }, { "epoch": 0.526158027257104, "grad_norm": 0.1642560064792633, "learning_rate": 9.645488767847042e-05, "loss": 0.0448, "step": 26330 }, { "epoch": 0.5263578593981056, "grad_norm": 0.1399822235107422, "learning_rate": 9.645106475031594e-05, "loss": 0.0431, "step": 26340 }, { "epoch": 0.5265576915391071, "grad_norm": 0.1443687379360199, "learning_rate": 9.644723983785729e-05, "loss": 0.0442, "step": 26350 }, { "epoch": 0.5267575236801088, "grad_norm": 0.15230752527713776, "learning_rate": 9.644341294125787e-05, "loss": 0.0408, "step": 26360 }, { "epoch": 0.5269573558211103, "grad_norm": 0.16752655804157257, "learning_rate": 9.643958406068118e-05, "loss": 0.036, "step": 26370 }, { "epoch": 0.5271571879621119, "grad_norm": 0.16669471561908722, "learning_rate": 9.643575319629079e-05, "loss": 0.039, "step": 26380 }, { "epoch": 0.5273570201031134, "grad_norm": 0.15311765670776367, "learning_rate": 9.64319203482503e-05, "loss": 0.0476, "step": 26390 }, { "epoch": 0.527556852244115, "grad_norm": 0.20661816000938416, "learning_rate": 9.642808551672348e-05, "loss": 0.056, "step": 26400 }, { "epoch": 0.5277566843851165, "grad_norm": 0.23677173256874084, "learning_rate": 9.642424870187414e-05, "loss": 0.041, "step": 26410 }, { "epoch": 0.5279565165261181, "grad_norm": 0.15403017401695251, "learning_rate": 9.642040990386618e-05, "loss": 0.0439, "step": 26420 }, { "epoch": 0.5281563486671196, "grad_norm": 0.12673978507518768, "learning_rate": 9.641656912286359e-05, "loss": 0.044, "step": 26430 }, { "epoch": 0.5283561808081212, "grad_norm": 0.16472207009792328, "learning_rate": 9.641272635903043e-05, "loss": 0.0487, "step": 26440 }, { "epoch": 0.5285560129491227, "grad_norm": 0.1978832185268402, "learning_rate": 9.640888161253087e-05, "loss": 0.0491, "step": 26450 }, { "epoch": 0.5287558450901243, "grad_norm": 0.1706671267747879, "learning_rate": 9.640503488352913e-05, "loss": 0.0448, "step": 26460 }, { "epoch": 0.5289556772311259, "grad_norm": 0.13422122597694397, "learning_rate": 9.640118617218956e-05, "loss": 0.0503, "step": 26470 }, { "epoch": 0.5291555093721274, "grad_norm": 0.14818644523620605, "learning_rate": 9.639733547867656e-05, "loss": 0.0367, "step": 26480 }, { "epoch": 0.529355341513129, "grad_norm": 0.12991417944431305, "learning_rate": 9.639348280315462e-05, "loss": 0.0496, "step": 26490 }, { "epoch": 0.5295551736541305, "grad_norm": 0.12841591238975525, "learning_rate": 9.638962814578832e-05, "loss": 0.0414, "step": 26500 }, { "epoch": 0.5297550057951321, "grad_norm": 0.14780232310295105, "learning_rate": 9.638577150674231e-05, "loss": 0.0449, "step": 26510 }, { "epoch": 0.5299548379361336, "grad_norm": 0.13445791602134705, "learning_rate": 9.638191288618139e-05, "loss": 0.0468, "step": 26520 }, { "epoch": 0.5301546700771352, "grad_norm": 0.16118645668029785, "learning_rate": 9.637805228427033e-05, "loss": 0.0388, "step": 26530 }, { "epoch": 0.5303545022181367, "grad_norm": 0.1614958792924881, "learning_rate": 9.637418970117409e-05, "loss": 0.0574, "step": 26540 }, { "epoch": 0.5305543343591383, "grad_norm": 0.12981493771076202, "learning_rate": 9.637032513705763e-05, "loss": 0.0458, "step": 26550 }, { "epoch": 0.5307541665001398, "grad_norm": 0.1669105589389801, "learning_rate": 9.63664585920861e-05, "loss": 0.038, "step": 26560 }, { "epoch": 0.5309539986411415, "grad_norm": 0.14216309785842896, "learning_rate": 9.636259006642463e-05, "loss": 0.0415, "step": 26570 }, { "epoch": 0.531153830782143, "grad_norm": 0.1420004814863205, "learning_rate": 9.635871956023848e-05, "loss": 0.049, "step": 26580 }, { "epoch": 0.5313536629231446, "grad_norm": 0.18139144778251648, "learning_rate": 9.635484707369297e-05, "loss": 0.0528, "step": 26590 }, { "epoch": 0.5315534950641461, "grad_norm": 0.18185092508792877, "learning_rate": 9.635097260695358e-05, "loss": 0.0462, "step": 26600 }, { "epoch": 0.5317533272051477, "grad_norm": 0.20917269587516785, "learning_rate": 9.634709616018578e-05, "loss": 0.0392, "step": 26610 }, { "epoch": 0.5319531593461493, "grad_norm": 0.172255277633667, "learning_rate": 9.634321773355518e-05, "loss": 0.0374, "step": 26620 }, { "epoch": 0.5321529914871508, "grad_norm": 0.14229890704154968, "learning_rate": 9.633933732722746e-05, "loss": 0.0394, "step": 26630 }, { "epoch": 0.5323528236281524, "grad_norm": 0.16570226848125458, "learning_rate": 9.633545494136836e-05, "loss": 0.0454, "step": 26640 }, { "epoch": 0.5325526557691539, "grad_norm": 0.1297590732574463, "learning_rate": 9.633157057614374e-05, "loss": 0.0403, "step": 26650 }, { "epoch": 0.5327524879101555, "grad_norm": 0.1502102166414261, "learning_rate": 9.632768423171956e-05, "loss": 0.0541, "step": 26660 }, { "epoch": 0.532952320051157, "grad_norm": 0.16999012231826782, "learning_rate": 9.632379590826182e-05, "loss": 0.0454, "step": 26670 }, { "epoch": 0.5331521521921586, "grad_norm": 0.13275843858718872, "learning_rate": 9.63199056059366e-05, "loss": 0.0369, "step": 26680 }, { "epoch": 0.5333519843331601, "grad_norm": 0.14630505442619324, "learning_rate": 9.631601332491011e-05, "loss": 0.0381, "step": 26690 }, { "epoch": 0.5335518164741617, "grad_norm": 0.11120978742837906, "learning_rate": 9.63121190653486e-05, "loss": 0.0367, "step": 26700 }, { "epoch": 0.5337516486151632, "grad_norm": 0.12593024969100952, "learning_rate": 9.630822282741848e-05, "loss": 0.0415, "step": 26710 }, { "epoch": 0.5339514807561648, "grad_norm": 0.14068175852298737, "learning_rate": 9.630432461128611e-05, "loss": 0.0411, "step": 26720 }, { "epoch": 0.5341513128971663, "grad_norm": 0.1541888564825058, "learning_rate": 9.630042441711809e-05, "loss": 0.0415, "step": 26730 }, { "epoch": 0.5343511450381679, "grad_norm": 0.14175944030284882, "learning_rate": 9.629652224508098e-05, "loss": 0.0349, "step": 26740 }, { "epoch": 0.5345509771791696, "grad_norm": 0.14508719742298126, "learning_rate": 9.629261809534149e-05, "loss": 0.0392, "step": 26750 }, { "epoch": 0.534750809320171, "grad_norm": 0.1425241231918335, "learning_rate": 9.62887119680664e-05, "loss": 0.0431, "step": 26760 }, { "epoch": 0.5349506414611727, "grad_norm": 0.13004645705223083, "learning_rate": 9.628480386342255e-05, "loss": 0.0406, "step": 26770 }, { "epoch": 0.5351504736021742, "grad_norm": 0.1368190050125122, "learning_rate": 9.628089378157694e-05, "loss": 0.0393, "step": 26780 }, { "epoch": 0.5353503057431758, "grad_norm": 0.14252014458179474, "learning_rate": 9.627698172269654e-05, "loss": 0.0369, "step": 26790 }, { "epoch": 0.5355501378841773, "grad_norm": 0.13264013826847076, "learning_rate": 9.627306768694853e-05, "loss": 0.0401, "step": 26800 }, { "epoch": 0.5357499700251789, "grad_norm": 0.1558779925107956, "learning_rate": 9.626915167450004e-05, "loss": 0.0438, "step": 26810 }, { "epoch": 0.5359498021661804, "grad_norm": 0.17479358613491058, "learning_rate": 9.62652336855184e-05, "loss": 0.0417, "step": 26820 }, { "epoch": 0.536149634307182, "grad_norm": 0.14231862127780914, "learning_rate": 9.626131372017098e-05, "loss": 0.0445, "step": 26830 }, { "epoch": 0.5363494664481835, "grad_norm": 0.12894438207149506, "learning_rate": 9.62573917786252e-05, "loss": 0.0458, "step": 26840 }, { "epoch": 0.5365492985891851, "grad_norm": 0.14121697843074799, "learning_rate": 9.625346786104864e-05, "loss": 0.0468, "step": 26850 }, { "epoch": 0.5367491307301866, "grad_norm": 0.1404270976781845, "learning_rate": 9.624954196760889e-05, "loss": 0.0441, "step": 26860 }, { "epoch": 0.5369489628711882, "grad_norm": 0.13151931762695312, "learning_rate": 9.624561409847368e-05, "loss": 0.048, "step": 26870 }, { "epoch": 0.5371487950121897, "grad_norm": 0.11968757957220078, "learning_rate": 9.624168425381079e-05, "loss": 0.0433, "step": 26880 }, { "epoch": 0.5373486271531913, "grad_norm": 0.14677506685256958, "learning_rate": 9.62377524337881e-05, "loss": 0.0421, "step": 26890 }, { "epoch": 0.5375484592941929, "grad_norm": 0.165362149477005, "learning_rate": 9.623381863857357e-05, "loss": 0.0442, "step": 26900 }, { "epoch": 0.5377482914351944, "grad_norm": 0.18144859373569489, "learning_rate": 9.622988286833523e-05, "loss": 0.0467, "step": 26910 }, { "epoch": 0.537948123576196, "grad_norm": 0.16080376505851746, "learning_rate": 9.622594512324123e-05, "loss": 0.0396, "step": 26920 }, { "epoch": 0.5381479557171975, "grad_norm": 0.11072314530611038, "learning_rate": 9.622200540345977e-05, "loss": 0.0398, "step": 26930 }, { "epoch": 0.5383477878581991, "grad_norm": 0.11951000243425369, "learning_rate": 9.621806370915916e-05, "loss": 0.0462, "step": 26940 }, { "epoch": 0.5385476199992006, "grad_norm": 0.10732541233301163, "learning_rate": 9.621412004050776e-05, "loss": 0.0343, "step": 26950 }, { "epoch": 0.5387474521402023, "grad_norm": 0.15554362535476685, "learning_rate": 9.621017439767405e-05, "loss": 0.042, "step": 26960 }, { "epoch": 0.5389472842812038, "grad_norm": 0.12495683878660202, "learning_rate": 9.620622678082659e-05, "loss": 0.0555, "step": 26970 }, { "epoch": 0.5391471164222054, "grad_norm": 0.1476251631975174, "learning_rate": 9.6202277190134e-05, "loss": 0.0476, "step": 26980 }, { "epoch": 0.5393469485632069, "grad_norm": 0.1253460943698883, "learning_rate": 9.6198325625765e-05, "loss": 0.0427, "step": 26990 }, { "epoch": 0.5395467807042085, "grad_norm": 0.16378942131996155, "learning_rate": 9.619437208788842e-05, "loss": 0.0431, "step": 27000 }, { "epoch": 0.53974661284521, "grad_norm": 0.17783918976783752, "learning_rate": 9.619041657667311e-05, "loss": 0.0444, "step": 27010 }, { "epoch": 0.5399464449862116, "grad_norm": 0.1521102637052536, "learning_rate": 9.618645909228806e-05, "loss": 0.0516, "step": 27020 }, { "epoch": 0.5401462771272132, "grad_norm": 0.152583047747612, "learning_rate": 9.618249963490233e-05, "loss": 0.0372, "step": 27030 }, { "epoch": 0.5403461092682147, "grad_norm": 0.1635286509990692, "learning_rate": 9.617853820468506e-05, "loss": 0.0333, "step": 27040 }, { "epoch": 0.5405459414092163, "grad_norm": 0.14121294021606445, "learning_rate": 9.617457480180545e-05, "loss": 0.037, "step": 27050 }, { "epoch": 0.5407457735502178, "grad_norm": 0.1343817412853241, "learning_rate": 9.617060942643283e-05, "loss": 0.0419, "step": 27060 }, { "epoch": 0.5409456056912194, "grad_norm": 0.14614267647266388, "learning_rate": 9.616664207873662e-05, "loss": 0.0467, "step": 27070 }, { "epoch": 0.5411454378322209, "grad_norm": 0.1614243984222412, "learning_rate": 9.616267275888624e-05, "loss": 0.0474, "step": 27080 }, { "epoch": 0.5413452699732225, "grad_norm": 0.16868077218532562, "learning_rate": 9.61587014670513e-05, "loss": 0.042, "step": 27090 }, { "epoch": 0.541545102114224, "grad_norm": 0.13408422470092773, "learning_rate": 9.615472820340143e-05, "loss": 0.0435, "step": 27100 }, { "epoch": 0.5417449342552256, "grad_norm": 0.16419930756092072, "learning_rate": 9.615075296810636e-05, "loss": 0.0439, "step": 27110 }, { "epoch": 0.5419447663962271, "grad_norm": 0.17445480823516846, "learning_rate": 9.61467757613359e-05, "loss": 0.0443, "step": 27120 }, { "epoch": 0.5421445985372287, "grad_norm": 0.10471932590007782, "learning_rate": 9.614279658325997e-05, "loss": 0.0316, "step": 27130 }, { "epoch": 0.5423444306782302, "grad_norm": 0.14003075659275055, "learning_rate": 9.613881543404852e-05, "loss": 0.0392, "step": 27140 }, { "epoch": 0.5425442628192318, "grad_norm": 0.1416909098625183, "learning_rate": 9.613483231387163e-05, "loss": 0.0482, "step": 27150 }, { "epoch": 0.5427440949602335, "grad_norm": 0.15264108777046204, "learning_rate": 9.613084722289946e-05, "loss": 0.0449, "step": 27160 }, { "epoch": 0.542943927101235, "grad_norm": 0.13158215582370758, "learning_rate": 9.612686016130226e-05, "loss": 0.0437, "step": 27170 }, { "epoch": 0.5431437592422366, "grad_norm": 0.10939313471317291, "learning_rate": 9.612287112925032e-05, "loss": 0.0358, "step": 27180 }, { "epoch": 0.5433435913832381, "grad_norm": 0.13707378506660461, "learning_rate": 9.611888012691405e-05, "loss": 0.0452, "step": 27190 }, { "epoch": 0.5435434235242397, "grad_norm": 0.12475236505270004, "learning_rate": 9.611488715446395e-05, "loss": 0.0386, "step": 27200 }, { "epoch": 0.5437432556652412, "grad_norm": 0.1476515233516693, "learning_rate": 9.611089221207058e-05, "loss": 0.0393, "step": 27210 }, { "epoch": 0.5439430878062428, "grad_norm": 0.13790607452392578, "learning_rate": 9.610689529990461e-05, "loss": 0.0413, "step": 27220 }, { "epoch": 0.5441429199472443, "grad_norm": 0.1732448935508728, "learning_rate": 9.610289641813676e-05, "loss": 0.0479, "step": 27230 }, { "epoch": 0.5443427520882459, "grad_norm": 0.136357843875885, "learning_rate": 9.609889556693788e-05, "loss": 0.0388, "step": 27240 }, { "epoch": 0.5445425842292474, "grad_norm": 0.1371600329875946, "learning_rate": 9.609489274647887e-05, "loss": 0.049, "step": 27250 }, { "epoch": 0.544742416370249, "grad_norm": 0.1736580729484558, "learning_rate": 9.609088795693073e-05, "loss": 0.0434, "step": 27260 }, { "epoch": 0.5449422485112505, "grad_norm": 0.12954026460647583, "learning_rate": 9.608688119846451e-05, "loss": 0.0394, "step": 27270 }, { "epoch": 0.5451420806522521, "grad_norm": 0.17117510735988617, "learning_rate": 9.608287247125139e-05, "loss": 0.0432, "step": 27280 }, { "epoch": 0.5453419127932536, "grad_norm": 0.1277005672454834, "learning_rate": 9.607886177546262e-05, "loss": 0.0417, "step": 27290 }, { "epoch": 0.5455417449342552, "grad_norm": 0.13665612041950226, "learning_rate": 9.607484911126954e-05, "loss": 0.0364, "step": 27300 }, { "epoch": 0.5457415770752568, "grad_norm": 0.12030945718288422, "learning_rate": 9.607083447884354e-05, "loss": 0.0528, "step": 27310 }, { "epoch": 0.5459414092162583, "grad_norm": 0.1424618810415268, "learning_rate": 9.60668178783561e-05, "loss": 0.0393, "step": 27320 }, { "epoch": 0.5461412413572599, "grad_norm": 0.15684367716312408, "learning_rate": 9.606279930997886e-05, "loss": 0.0397, "step": 27330 }, { "epoch": 0.5463410734982614, "grad_norm": 0.1398071050643921, "learning_rate": 9.605877877388343e-05, "loss": 0.0343, "step": 27340 }, { "epoch": 0.546540905639263, "grad_norm": 0.1328730434179306, "learning_rate": 9.60547562702416e-05, "loss": 0.0476, "step": 27350 }, { "epoch": 0.5467407377802646, "grad_norm": 0.15517926216125488, "learning_rate": 9.605073179922518e-05, "loss": 0.0458, "step": 27360 }, { "epoch": 0.5469405699212662, "grad_norm": 0.14678534865379333, "learning_rate": 9.60467053610061e-05, "loss": 0.044, "step": 27370 }, { "epoch": 0.5471404020622677, "grad_norm": 0.15046250820159912, "learning_rate": 9.604267695575634e-05, "loss": 0.0384, "step": 27380 }, { "epoch": 0.5473402342032693, "grad_norm": 0.14541876316070557, "learning_rate": 9.603864658364802e-05, "loss": 0.0475, "step": 27390 }, { "epoch": 0.5475400663442708, "grad_norm": 0.16166406869888306, "learning_rate": 9.60346142448533e-05, "loss": 0.039, "step": 27400 }, { "epoch": 0.5477398984852724, "grad_norm": 0.1821003556251526, "learning_rate": 9.603057993954443e-05, "loss": 0.0458, "step": 27410 }, { "epoch": 0.5479397306262739, "grad_norm": 0.1912255436182022, "learning_rate": 9.602654366789374e-05, "loss": 0.0467, "step": 27420 }, { "epoch": 0.5481395627672755, "grad_norm": 0.14936862885951996, "learning_rate": 9.602250543007364e-05, "loss": 0.0325, "step": 27430 }, { "epoch": 0.5483393949082771, "grad_norm": 0.19219724833965302, "learning_rate": 9.601846522625668e-05, "loss": 0.0426, "step": 27440 }, { "epoch": 0.5485392270492786, "grad_norm": 0.20025108754634857, "learning_rate": 9.601442305661541e-05, "loss": 0.0427, "step": 27450 }, { "epoch": 0.5487390591902802, "grad_norm": 0.20578047633171082, "learning_rate": 9.601037892132253e-05, "loss": 0.0435, "step": 27460 }, { "epoch": 0.5489388913312817, "grad_norm": 0.12416473031044006, "learning_rate": 9.600633282055076e-05, "loss": 0.0318, "step": 27470 }, { "epoch": 0.5491387234722833, "grad_norm": 0.15802593529224396, "learning_rate": 9.600228475447298e-05, "loss": 0.0509, "step": 27480 }, { "epoch": 0.5493385556132848, "grad_norm": 0.1582522839307785, "learning_rate": 9.599823472326212e-05, "loss": 0.0422, "step": 27490 }, { "epoch": 0.5495383877542864, "grad_norm": 0.1726619303226471, "learning_rate": 9.599418272709115e-05, "loss": 0.0489, "step": 27500 }, { "epoch": 0.5497382198952879, "grad_norm": 0.15857532620429993, "learning_rate": 9.599012876613321e-05, "loss": 0.0386, "step": 27510 }, { "epoch": 0.5499380520362895, "grad_norm": 0.12278904765844345, "learning_rate": 9.598607284056145e-05, "loss": 0.043, "step": 27520 }, { "epoch": 0.550137884177291, "grad_norm": 0.16907894611358643, "learning_rate": 9.598201495054913e-05, "loss": 0.0412, "step": 27530 }, { "epoch": 0.5503377163182926, "grad_norm": 0.14068186283111572, "learning_rate": 9.59779550962696e-05, "loss": 0.0438, "step": 27540 }, { "epoch": 0.5505375484592941, "grad_norm": 0.1214575245976448, "learning_rate": 9.59738932778963e-05, "loss": 0.0424, "step": 27550 }, { "epoch": 0.5507373806002958, "grad_norm": 0.1705930083990097, "learning_rate": 9.596982949560274e-05, "loss": 0.0389, "step": 27560 }, { "epoch": 0.5509372127412973, "grad_norm": 0.1351352035999298, "learning_rate": 9.59657637495625e-05, "loss": 0.0478, "step": 27570 }, { "epoch": 0.5511370448822989, "grad_norm": 0.1378231644630432, "learning_rate": 9.59616960399493e-05, "loss": 0.0424, "step": 27580 }, { "epoch": 0.5513368770233005, "grad_norm": 0.15883438289165497, "learning_rate": 9.595762636693685e-05, "loss": 0.0478, "step": 27590 }, { "epoch": 0.551536709164302, "grad_norm": 0.1951666623353958, "learning_rate": 9.595355473069905e-05, "loss": 0.0398, "step": 27600 }, { "epoch": 0.5517365413053036, "grad_norm": 0.17038963735103607, "learning_rate": 9.594948113140979e-05, "loss": 0.0496, "step": 27610 }, { "epoch": 0.5519363734463051, "grad_norm": 0.12292885780334473, "learning_rate": 9.594540556924312e-05, "loss": 0.0459, "step": 27620 }, { "epoch": 0.5521362055873067, "grad_norm": 0.1276852786540985, "learning_rate": 9.594132804437313e-05, "loss": 0.0362, "step": 27630 }, { "epoch": 0.5523360377283082, "grad_norm": 0.15220317244529724, "learning_rate": 9.593724855697401e-05, "loss": 0.0503, "step": 27640 }, { "epoch": 0.5525358698693098, "grad_norm": 0.13863442838191986, "learning_rate": 9.593316710722001e-05, "loss": 0.0417, "step": 27650 }, { "epoch": 0.5527357020103113, "grad_norm": 0.12900540232658386, "learning_rate": 9.592908369528548e-05, "loss": 0.0414, "step": 27660 }, { "epoch": 0.5529355341513129, "grad_norm": 0.1500018835067749, "learning_rate": 9.592499832134489e-05, "loss": 0.0452, "step": 27670 }, { "epoch": 0.5531353662923144, "grad_norm": 0.11991036683320999, "learning_rate": 9.592091098557275e-05, "loss": 0.0399, "step": 27680 }, { "epoch": 0.553335198433316, "grad_norm": 0.15205661952495575, "learning_rate": 9.591682168814363e-05, "loss": 0.0395, "step": 27690 }, { "epoch": 0.5535350305743175, "grad_norm": 0.16498591005802155, "learning_rate": 9.591273042923223e-05, "loss": 0.0367, "step": 27700 }, { "epoch": 0.5537348627153191, "grad_norm": 0.11401980370283127, "learning_rate": 9.590863720901336e-05, "loss": 0.0454, "step": 27710 }, { "epoch": 0.5539346948563207, "grad_norm": 0.1262691169977188, "learning_rate": 9.590454202766182e-05, "loss": 0.0441, "step": 27720 }, { "epoch": 0.5541345269973222, "grad_norm": 0.17358136177062988, "learning_rate": 9.59004448853526e-05, "loss": 0.0368, "step": 27730 }, { "epoch": 0.5543343591383239, "grad_norm": 0.1311301738023758, "learning_rate": 9.589634578226067e-05, "loss": 0.0368, "step": 27740 }, { "epoch": 0.5545341912793254, "grad_norm": 0.1669553518295288, "learning_rate": 9.589224471856118e-05, "loss": 0.0398, "step": 27750 }, { "epoch": 0.554734023420327, "grad_norm": 0.160566046833992, "learning_rate": 9.588814169442928e-05, "loss": 0.0379, "step": 27760 }, { "epoch": 0.5549338555613285, "grad_norm": 0.16270890831947327, "learning_rate": 9.588403671004029e-05, "loss": 0.0439, "step": 27770 }, { "epoch": 0.5551336877023301, "grad_norm": 0.14215604960918427, "learning_rate": 9.587992976556953e-05, "loss": 0.0459, "step": 27780 }, { "epoch": 0.5553335198433316, "grad_norm": 0.1730937510728836, "learning_rate": 9.587582086119247e-05, "loss": 0.0446, "step": 27790 }, { "epoch": 0.5555333519843332, "grad_norm": 0.13632670044898987, "learning_rate": 9.58717099970846e-05, "loss": 0.0483, "step": 27800 }, { "epoch": 0.5557331841253347, "grad_norm": 0.11995267122983932, "learning_rate": 9.586759717342155e-05, "loss": 0.0341, "step": 27810 }, { "epoch": 0.5559330162663363, "grad_norm": 0.1416911780834198, "learning_rate": 9.586348239037903e-05, "loss": 0.0445, "step": 27820 }, { "epoch": 0.5561328484073378, "grad_norm": 0.16091814637184143, "learning_rate": 9.585936564813278e-05, "loss": 0.0446, "step": 27830 }, { "epoch": 0.5563326805483394, "grad_norm": 0.11872809380292892, "learning_rate": 9.585524694685868e-05, "loss": 0.0492, "step": 27840 }, { "epoch": 0.556532512689341, "grad_norm": 0.12954236567020416, "learning_rate": 9.585112628673266e-05, "loss": 0.0415, "step": 27850 }, { "epoch": 0.5567323448303425, "grad_norm": 0.15553700923919678, "learning_rate": 9.584700366793077e-05, "loss": 0.0354, "step": 27860 }, { "epoch": 0.5569321769713441, "grad_norm": 0.16543526947498322, "learning_rate": 9.58428790906291e-05, "loss": 0.0439, "step": 27870 }, { "epoch": 0.5571320091123456, "grad_norm": 0.14672833681106567, "learning_rate": 9.583875255500385e-05, "loss": 0.037, "step": 27880 }, { "epoch": 0.5573318412533472, "grad_norm": 0.1660987138748169, "learning_rate": 9.58346240612313e-05, "loss": 0.0452, "step": 27890 }, { "epoch": 0.5575316733943487, "grad_norm": 0.15586426854133606, "learning_rate": 9.583049360948782e-05, "loss": 0.0503, "step": 27900 }, { "epoch": 0.5577315055353503, "grad_norm": 0.11644305288791656, "learning_rate": 9.582636119994983e-05, "loss": 0.0382, "step": 27910 }, { "epoch": 0.5579313376763518, "grad_norm": 0.14447113871574402, "learning_rate": 9.58222268327939e-05, "loss": 0.0428, "step": 27920 }, { "epoch": 0.5581311698173534, "grad_norm": 0.13684840500354767, "learning_rate": 9.58180905081966e-05, "loss": 0.037, "step": 27930 }, { "epoch": 0.5583310019583549, "grad_norm": 0.14132221043109894, "learning_rate": 9.581395222633463e-05, "loss": 0.0432, "step": 27940 }, { "epoch": 0.5585308340993566, "grad_norm": 0.15681743621826172, "learning_rate": 9.580981198738482e-05, "loss": 0.0421, "step": 27950 }, { "epoch": 0.558730666240358, "grad_norm": 0.11545481532812119, "learning_rate": 9.580566979152397e-05, "loss": 0.0383, "step": 27960 }, { "epoch": 0.5589304983813597, "grad_norm": 0.1548442244529724, "learning_rate": 9.580152563892906e-05, "loss": 0.0442, "step": 27970 }, { "epoch": 0.5591303305223612, "grad_norm": 0.14962472021579742, "learning_rate": 9.579737952977712e-05, "loss": 0.0396, "step": 27980 }, { "epoch": 0.5593301626633628, "grad_norm": 0.14360854029655457, "learning_rate": 9.579323146424525e-05, "loss": 0.0402, "step": 27990 }, { "epoch": 0.5595299948043644, "grad_norm": 0.11739914119243622, "learning_rate": 9.578908144251067e-05, "loss": 0.0348, "step": 28000 }, { "epoch": 0.5597298269453659, "grad_norm": 0.16357463598251343, "learning_rate": 9.578492946475064e-05, "loss": 0.0416, "step": 28010 }, { "epoch": 0.5599296590863675, "grad_norm": 0.1893538236618042, "learning_rate": 9.578077553114253e-05, "loss": 0.0411, "step": 28020 }, { "epoch": 0.560129491227369, "grad_norm": 0.1855582445859909, "learning_rate": 9.577661964186378e-05, "loss": 0.0444, "step": 28030 }, { "epoch": 0.5603293233683706, "grad_norm": 0.14543616771697998, "learning_rate": 9.577246179709196e-05, "loss": 0.047, "step": 28040 }, { "epoch": 0.5605291555093721, "grad_norm": 0.16075675189495087, "learning_rate": 9.576830199700464e-05, "loss": 0.0398, "step": 28050 }, { "epoch": 0.5607289876503737, "grad_norm": 0.13079805672168732, "learning_rate": 9.576414024177956e-05, "loss": 0.0406, "step": 28060 }, { "epoch": 0.5609288197913752, "grad_norm": 0.12056397646665573, "learning_rate": 9.575997653159445e-05, "loss": 0.0435, "step": 28070 }, { "epoch": 0.5611286519323768, "grad_norm": 0.13283953070640564, "learning_rate": 9.575581086662721e-05, "loss": 0.0376, "step": 28080 }, { "epoch": 0.5613284840733783, "grad_norm": 0.14269553124904633, "learning_rate": 9.575164324705581e-05, "loss": 0.0344, "step": 28090 }, { "epoch": 0.5615283162143799, "grad_norm": 0.15233691036701202, "learning_rate": 9.574747367305825e-05, "loss": 0.0377, "step": 28100 }, { "epoch": 0.5617281483553814, "grad_norm": 0.13275015354156494, "learning_rate": 9.574330214481264e-05, "loss": 0.041, "step": 28110 }, { "epoch": 0.561927980496383, "grad_norm": 0.17141954600811005, "learning_rate": 9.573912866249721e-05, "loss": 0.0455, "step": 28120 }, { "epoch": 0.5621278126373846, "grad_norm": 0.17252036929130554, "learning_rate": 9.573495322629022e-05, "loss": 0.0391, "step": 28130 }, { "epoch": 0.5623276447783861, "grad_norm": 0.14419247210025787, "learning_rate": 9.573077583637006e-05, "loss": 0.0477, "step": 28140 }, { "epoch": 0.5625274769193878, "grad_norm": 0.17640109360218048, "learning_rate": 9.572659649291516e-05, "loss": 0.0394, "step": 28150 }, { "epoch": 0.5627273090603893, "grad_norm": 0.13338473439216614, "learning_rate": 9.572241519610405e-05, "loss": 0.0387, "step": 28160 }, { "epoch": 0.5629271412013909, "grad_norm": 0.138657808303833, "learning_rate": 9.571823194611537e-05, "loss": 0.0379, "step": 28170 }, { "epoch": 0.5631269733423924, "grad_norm": 0.14278525114059448, "learning_rate": 9.57140467431278e-05, "loss": 0.0451, "step": 28180 }, { "epoch": 0.563326805483394, "grad_norm": 0.1357237845659256, "learning_rate": 9.570985958732013e-05, "loss": 0.0339, "step": 28190 }, { "epoch": 0.5635266376243955, "grad_norm": 0.12859830260276794, "learning_rate": 9.570567047887124e-05, "loss": 0.0462, "step": 28200 }, { "epoch": 0.5637264697653971, "grad_norm": 0.14837025105953217, "learning_rate": 9.570147941796007e-05, "loss": 0.0349, "step": 28210 }, { "epoch": 0.5639263019063986, "grad_norm": 0.15502406656742096, "learning_rate": 9.569728640476566e-05, "loss": 0.037, "step": 28220 }, { "epoch": 0.5641261340474002, "grad_norm": 0.18129843473434448, "learning_rate": 9.569309143946712e-05, "loss": 0.0391, "step": 28230 }, { "epoch": 0.5643259661884017, "grad_norm": 0.1210358515381813, "learning_rate": 9.568889452224364e-05, "loss": 0.0412, "step": 28240 }, { "epoch": 0.5645257983294033, "grad_norm": 0.13652177155017853, "learning_rate": 9.568469565327454e-05, "loss": 0.0437, "step": 28250 }, { "epoch": 0.5647256304704048, "grad_norm": 0.12883692979812622, "learning_rate": 9.568049483273915e-05, "loss": 0.0445, "step": 28260 }, { "epoch": 0.5649254626114064, "grad_norm": 0.17057573795318604, "learning_rate": 9.567629206081695e-05, "loss": 0.0468, "step": 28270 }, { "epoch": 0.565125294752408, "grad_norm": 0.12160658836364746, "learning_rate": 9.567208733768746e-05, "loss": 0.0435, "step": 28280 }, { "epoch": 0.5653251268934095, "grad_norm": 0.1224149763584137, "learning_rate": 9.566788066353029e-05, "loss": 0.041, "step": 28290 }, { "epoch": 0.5655249590344111, "grad_norm": 0.12346886098384857, "learning_rate": 9.566367203852518e-05, "loss": 0.0367, "step": 28300 }, { "epoch": 0.5657247911754126, "grad_norm": 0.12779922783374786, "learning_rate": 9.565946146285186e-05, "loss": 0.037, "step": 28310 }, { "epoch": 0.5659246233164142, "grad_norm": 0.16530460119247437, "learning_rate": 9.565524893669024e-05, "loss": 0.0384, "step": 28320 }, { "epoch": 0.5661244554574157, "grad_norm": 0.39531973004341125, "learning_rate": 9.565103446022025e-05, "loss": 0.0432, "step": 28330 }, { "epoch": 0.5663242875984174, "grad_norm": 0.14156147837638855, "learning_rate": 9.564681803362193e-05, "loss": 0.0376, "step": 28340 }, { "epoch": 0.5665241197394189, "grad_norm": 0.18519562482833862, "learning_rate": 9.564259965707541e-05, "loss": 0.047, "step": 28350 }, { "epoch": 0.5667239518804205, "grad_norm": 0.16468922793865204, "learning_rate": 9.563837933076088e-05, "loss": 0.0374, "step": 28360 }, { "epoch": 0.566923784021422, "grad_norm": 0.16164137423038483, "learning_rate": 9.56341570548586e-05, "loss": 0.0385, "step": 28370 }, { "epoch": 0.5671236161624236, "grad_norm": 0.12433762103319168, "learning_rate": 9.5629932829549e-05, "loss": 0.0398, "step": 28380 }, { "epoch": 0.5673234483034251, "grad_norm": 0.13374687731266022, "learning_rate": 9.562570665501247e-05, "loss": 0.039, "step": 28390 }, { "epoch": 0.5675232804444267, "grad_norm": 0.1525164395570755, "learning_rate": 9.562147853142958e-05, "loss": 0.0415, "step": 28400 }, { "epoch": 0.5677231125854283, "grad_norm": 0.127887561917305, "learning_rate": 9.561724845898092e-05, "loss": 0.0435, "step": 28410 }, { "epoch": 0.5679229447264298, "grad_norm": 0.13354849815368652, "learning_rate": 9.561301643784724e-05, "loss": 0.041, "step": 28420 }, { "epoch": 0.5681227768674314, "grad_norm": 0.1510118693113327, "learning_rate": 9.560878246820925e-05, "loss": 0.0501, "step": 28430 }, { "epoch": 0.5683226090084329, "grad_norm": 0.13559995591640472, "learning_rate": 9.560454655024789e-05, "loss": 0.035, "step": 28440 }, { "epoch": 0.5685224411494345, "grad_norm": 0.11343181133270264, "learning_rate": 9.560030868414408e-05, "loss": 0.0437, "step": 28450 }, { "epoch": 0.568722273290436, "grad_norm": 0.12988178431987762, "learning_rate": 9.559606887007884e-05, "loss": 0.0478, "step": 28460 }, { "epoch": 0.5689221054314376, "grad_norm": 0.15049484372138977, "learning_rate": 9.55918271082333e-05, "loss": 0.0369, "step": 28470 }, { "epoch": 0.5691219375724391, "grad_norm": 0.12806569039821625, "learning_rate": 9.558758339878866e-05, "loss": 0.0373, "step": 28480 }, { "epoch": 0.5693217697134407, "grad_norm": 0.1208944246172905, "learning_rate": 9.558333774192622e-05, "loss": 0.0449, "step": 28490 }, { "epoch": 0.5695216018544422, "grad_norm": 0.10979951918125153, "learning_rate": 9.557909013782731e-05, "loss": 0.043, "step": 28500 }, { "epoch": 0.5697214339954438, "grad_norm": 0.14095541834831238, "learning_rate": 9.557484058667342e-05, "loss": 0.0487, "step": 28510 }, { "epoch": 0.5699212661364453, "grad_norm": 0.13248538970947266, "learning_rate": 9.557058908864606e-05, "loss": 0.0359, "step": 28520 }, { "epoch": 0.570121098277447, "grad_norm": 0.16858617961406708, "learning_rate": 9.556633564392684e-05, "loss": 0.0404, "step": 28530 }, { "epoch": 0.5703209304184486, "grad_norm": 0.14089632034301758, "learning_rate": 9.556208025269748e-05, "loss": 0.0364, "step": 28540 }, { "epoch": 0.5705207625594501, "grad_norm": 0.10901785641908646, "learning_rate": 9.555782291513974e-05, "loss": 0.0382, "step": 28550 }, { "epoch": 0.5707205947004517, "grad_norm": 0.16003109514713287, "learning_rate": 9.555356363143551e-05, "loss": 0.0421, "step": 28560 }, { "epoch": 0.5709204268414532, "grad_norm": 0.18099112808704376, "learning_rate": 9.554930240176672e-05, "loss": 0.0319, "step": 28570 }, { "epoch": 0.5711202589824548, "grad_norm": 0.11980145424604416, "learning_rate": 9.554503922631539e-05, "loss": 0.0328, "step": 28580 }, { "epoch": 0.5713200911234563, "grad_norm": 0.1103302612900734, "learning_rate": 9.554077410526368e-05, "loss": 0.0383, "step": 28590 }, { "epoch": 0.5715199232644579, "grad_norm": 0.11088954657316208, "learning_rate": 9.553650703879375e-05, "loss": 0.0418, "step": 28600 }, { "epoch": 0.5717197554054594, "grad_norm": 0.11697743088006973, "learning_rate": 9.55322380270879e-05, "loss": 0.0352, "step": 28610 }, { "epoch": 0.571919587546461, "grad_norm": 0.13849487900733948, "learning_rate": 9.552796707032848e-05, "loss": 0.0365, "step": 28620 }, { "epoch": 0.5721194196874625, "grad_norm": 0.11981812119483948, "learning_rate": 9.552369416869795e-05, "loss": 0.0538, "step": 28630 }, { "epoch": 0.5723192518284641, "grad_norm": 0.23110425472259521, "learning_rate": 9.551941932237883e-05, "loss": 0.0461, "step": 28640 }, { "epoch": 0.5725190839694656, "grad_norm": 0.13417597115039825, "learning_rate": 9.551514253155373e-05, "loss": 0.0456, "step": 28650 }, { "epoch": 0.5727189161104672, "grad_norm": 0.16179662942886353, "learning_rate": 9.551086379640537e-05, "loss": 0.0441, "step": 28660 }, { "epoch": 0.5729187482514687, "grad_norm": 0.1527184098958969, "learning_rate": 9.550658311711651e-05, "loss": 0.038, "step": 28670 }, { "epoch": 0.5731185803924703, "grad_norm": 0.17325320839881897, "learning_rate": 9.550230049387002e-05, "loss": 0.0502, "step": 28680 }, { "epoch": 0.5733184125334719, "grad_norm": 0.1445750892162323, "learning_rate": 9.549801592684884e-05, "loss": 0.0415, "step": 28690 }, { "epoch": 0.5735182446744734, "grad_norm": 0.13545235991477966, "learning_rate": 9.5493729416236e-05, "loss": 0.041, "step": 28700 }, { "epoch": 0.573718076815475, "grad_norm": 0.17810679972171783, "learning_rate": 9.548944096221461e-05, "loss": 0.0432, "step": 28710 }, { "epoch": 0.5739179089564765, "grad_norm": 0.13592447340488434, "learning_rate": 9.548515056496787e-05, "loss": 0.0445, "step": 28720 }, { "epoch": 0.5741177410974782, "grad_norm": 0.15945640206336975, "learning_rate": 9.548085822467907e-05, "loss": 0.0425, "step": 28730 }, { "epoch": 0.5743175732384797, "grad_norm": 0.1628301739692688, "learning_rate": 9.547656394153155e-05, "loss": 0.0398, "step": 28740 }, { "epoch": 0.5745174053794813, "grad_norm": 0.16096985340118408, "learning_rate": 9.547226771570877e-05, "loss": 0.0435, "step": 28750 }, { "epoch": 0.5747172375204828, "grad_norm": 0.13716866075992584, "learning_rate": 9.546796954739423e-05, "loss": 0.0453, "step": 28760 }, { "epoch": 0.5749170696614844, "grad_norm": 0.16554321348667145, "learning_rate": 9.546366943677156e-05, "loss": 0.0461, "step": 28770 }, { "epoch": 0.5751169018024859, "grad_norm": 0.13177058100700378, "learning_rate": 9.545936738402446e-05, "loss": 0.0311, "step": 28780 }, { "epoch": 0.5753167339434875, "grad_norm": 0.19524294137954712, "learning_rate": 9.54550633893367e-05, "loss": 0.0385, "step": 28790 }, { "epoch": 0.575516566084489, "grad_norm": 0.1572098284959793, "learning_rate": 9.545075745289213e-05, "loss": 0.0443, "step": 28800 }, { "epoch": 0.5757163982254906, "grad_norm": 0.12959067523479462, "learning_rate": 9.544644957487469e-05, "loss": 0.0428, "step": 28810 }, { "epoch": 0.5759162303664922, "grad_norm": 0.1669958084821701, "learning_rate": 9.544213975546843e-05, "loss": 0.0458, "step": 28820 }, { "epoch": 0.5761160625074937, "grad_norm": 0.12201620638370514, "learning_rate": 9.543782799485743e-05, "loss": 0.0401, "step": 28830 }, { "epoch": 0.5763158946484953, "grad_norm": 0.11864753812551498, "learning_rate": 9.543351429322589e-05, "loss": 0.0375, "step": 28840 }, { "epoch": 0.5765157267894968, "grad_norm": 0.17235122621059418, "learning_rate": 9.542919865075807e-05, "loss": 0.0368, "step": 28850 }, { "epoch": 0.5767155589304984, "grad_norm": 0.13352656364440918, "learning_rate": 9.542488106763836e-05, "loss": 0.0347, "step": 28860 }, { "epoch": 0.5769153910714999, "grad_norm": 0.176670104265213, "learning_rate": 9.542056154405117e-05, "loss": 0.0372, "step": 28870 }, { "epoch": 0.5771152232125015, "grad_norm": 0.20112697780132294, "learning_rate": 9.541624008018105e-05, "loss": 0.0385, "step": 28880 }, { "epoch": 0.577315055353503, "grad_norm": 0.1712561696767807, "learning_rate": 9.541191667621257e-05, "loss": 0.0416, "step": 28890 }, { "epoch": 0.5775148874945046, "grad_norm": 0.16053327918052673, "learning_rate": 9.540759133233041e-05, "loss": 0.0508, "step": 28900 }, { "epoch": 0.5777147196355061, "grad_norm": 0.14591430127620697, "learning_rate": 9.54032640487194e-05, "loss": 0.039, "step": 28910 }, { "epoch": 0.5779145517765077, "grad_norm": 0.11470657587051392, "learning_rate": 9.539893482556434e-05, "loss": 0.0429, "step": 28920 }, { "epoch": 0.5781143839175092, "grad_norm": 0.1372375786304474, "learning_rate": 9.53946036630502e-05, "loss": 0.0486, "step": 28930 }, { "epoch": 0.5783142160585109, "grad_norm": 0.19394026696681976, "learning_rate": 9.539027056136198e-05, "loss": 0.0413, "step": 28940 }, { "epoch": 0.5785140481995125, "grad_norm": 0.19641099870204926, "learning_rate": 9.538593552068477e-05, "loss": 0.0433, "step": 28950 }, { "epoch": 0.578713880340514, "grad_norm": 0.12723499536514282, "learning_rate": 9.538159854120378e-05, "loss": 0.0462, "step": 28960 }, { "epoch": 0.5789137124815156, "grad_norm": 0.14430266618728638, "learning_rate": 9.537725962310427e-05, "loss": 0.0421, "step": 28970 }, { "epoch": 0.5791135446225171, "grad_norm": 0.16271616518497467, "learning_rate": 9.537291876657159e-05, "loss": 0.0426, "step": 28980 }, { "epoch": 0.5793133767635187, "grad_norm": 0.14446943998336792, "learning_rate": 9.536857597179119e-05, "loss": 0.036, "step": 28990 }, { "epoch": 0.5795132089045202, "grad_norm": 0.18986904621124268, "learning_rate": 9.536423123894854e-05, "loss": 0.0324, "step": 29000 }, { "epoch": 0.5797130410455218, "grad_norm": 0.15009751915931702, "learning_rate": 9.535988456822929e-05, "loss": 0.0422, "step": 29010 }, { "epoch": 0.5799128731865233, "grad_norm": 0.17120178043842316, "learning_rate": 9.535553595981909e-05, "loss": 0.0386, "step": 29020 }, { "epoch": 0.5801127053275249, "grad_norm": 0.13784588873386383, "learning_rate": 9.535118541390372e-05, "loss": 0.033, "step": 29030 }, { "epoch": 0.5803125374685264, "grad_norm": 0.14890927076339722, "learning_rate": 9.534683293066902e-05, "loss": 0.041, "step": 29040 }, { "epoch": 0.580512369609528, "grad_norm": 0.13921378552913666, "learning_rate": 9.534247851030092e-05, "loss": 0.0428, "step": 29050 }, { "epoch": 0.5807122017505295, "grad_norm": 0.14452403783798218, "learning_rate": 9.533812215298543e-05, "loss": 0.0392, "step": 29060 }, { "epoch": 0.5809120338915311, "grad_norm": 0.140745148062706, "learning_rate": 9.533376385890864e-05, "loss": 0.0469, "step": 29070 }, { "epoch": 0.5811118660325326, "grad_norm": 0.17110328376293182, "learning_rate": 9.532940362825676e-05, "loss": 0.0436, "step": 29080 }, { "epoch": 0.5813116981735342, "grad_norm": 0.18386763334274292, "learning_rate": 9.532504146121602e-05, "loss": 0.037, "step": 29090 }, { "epoch": 0.5815115303145358, "grad_norm": 0.18220844864845276, "learning_rate": 9.532067735797279e-05, "loss": 0.0469, "step": 29100 }, { "epoch": 0.5817113624555373, "grad_norm": 0.13072478771209717, "learning_rate": 9.531631131871346e-05, "loss": 0.043, "step": 29110 }, { "epoch": 0.581911194596539, "grad_norm": 0.1303601711988449, "learning_rate": 9.531194334362457e-05, "loss": 0.041, "step": 29120 }, { "epoch": 0.5821110267375404, "grad_norm": 0.19600914418697357, "learning_rate": 9.53075734328927e-05, "loss": 0.0439, "step": 29130 }, { "epoch": 0.5823108588785421, "grad_norm": 0.1353236883878708, "learning_rate": 9.530320158670452e-05, "loss": 0.0417, "step": 29140 }, { "epoch": 0.5825106910195436, "grad_norm": 0.15568697452545166, "learning_rate": 9.529882780524679e-05, "loss": 0.0462, "step": 29150 }, { "epoch": 0.5827105231605452, "grad_norm": 0.15749791264533997, "learning_rate": 9.529445208870635e-05, "loss": 0.0392, "step": 29160 }, { "epoch": 0.5829103553015467, "grad_norm": 0.13705679774284363, "learning_rate": 9.529007443727015e-05, "loss": 0.0419, "step": 29170 }, { "epoch": 0.5831101874425483, "grad_norm": 0.11841993033885956, "learning_rate": 9.528569485112514e-05, "loss": 0.0358, "step": 29180 }, { "epoch": 0.5833100195835498, "grad_norm": 0.18326382339000702, "learning_rate": 9.528131333045846e-05, "loss": 0.0438, "step": 29190 }, { "epoch": 0.5835098517245514, "grad_norm": 0.1477378010749817, "learning_rate": 9.527692987545726e-05, "loss": 0.0397, "step": 29200 }, { "epoch": 0.5837096838655529, "grad_norm": 0.2128623127937317, "learning_rate": 9.527254448630877e-05, "loss": 0.0411, "step": 29210 }, { "epoch": 0.5839095160065545, "grad_norm": 0.15647922456264496, "learning_rate": 9.526815716320036e-05, "loss": 0.0485, "step": 29220 }, { "epoch": 0.5841093481475561, "grad_norm": 0.1447426825761795, "learning_rate": 9.526376790631945e-05, "loss": 0.0367, "step": 29230 }, { "epoch": 0.5843091802885576, "grad_norm": 0.17965978384017944, "learning_rate": 9.525937671585352e-05, "loss": 0.041, "step": 29240 }, { "epoch": 0.5845090124295592, "grad_norm": 0.13812856376171112, "learning_rate": 9.525498359199014e-05, "loss": 0.0556, "step": 29250 }, { "epoch": 0.5847088445705607, "grad_norm": 0.18608656525611877, "learning_rate": 9.525058853491702e-05, "loss": 0.0433, "step": 29260 }, { "epoch": 0.5849086767115623, "grad_norm": 0.1312830150127411, "learning_rate": 9.524619154482189e-05, "loss": 0.0475, "step": 29270 }, { "epoch": 0.5851085088525638, "grad_norm": 0.1457752287387848, "learning_rate": 9.524179262189256e-05, "loss": 0.0426, "step": 29280 }, { "epoch": 0.5853083409935654, "grad_norm": 0.1430331915616989, "learning_rate": 9.523739176631696e-05, "loss": 0.0409, "step": 29290 }, { "epoch": 0.5855081731345669, "grad_norm": 0.09997048228979111, "learning_rate": 9.52329889782831e-05, "loss": 0.0374, "step": 29300 }, { "epoch": 0.5857080052755685, "grad_norm": 0.15363506972789764, "learning_rate": 9.522858425797905e-05, "loss": 0.0394, "step": 29310 }, { "epoch": 0.58590783741657, "grad_norm": 0.13469694554805756, "learning_rate": 9.522417760559295e-05, "loss": 0.0384, "step": 29320 }, { "epoch": 0.5861076695575717, "grad_norm": 0.12763150036334991, "learning_rate": 9.521976902131307e-05, "loss": 0.0379, "step": 29330 }, { "epoch": 0.5863075016985732, "grad_norm": 0.11518189311027527, "learning_rate": 9.521535850532774e-05, "loss": 0.0363, "step": 29340 }, { "epoch": 0.5865073338395748, "grad_norm": 0.1299910694360733, "learning_rate": 9.521094605782535e-05, "loss": 0.0423, "step": 29350 }, { "epoch": 0.5867071659805763, "grad_norm": 0.11773771792650223, "learning_rate": 9.520653167899438e-05, "loss": 0.0375, "step": 29360 }, { "epoch": 0.5869069981215779, "grad_norm": 0.12986066937446594, "learning_rate": 9.520211536902344e-05, "loss": 0.0357, "step": 29370 }, { "epoch": 0.5871068302625795, "grad_norm": 0.1166347786784172, "learning_rate": 9.519769712810118e-05, "loss": 0.0406, "step": 29380 }, { "epoch": 0.587306662403581, "grad_norm": 0.12529760599136353, "learning_rate": 9.519327695641632e-05, "loss": 0.0388, "step": 29390 }, { "epoch": 0.5875064945445826, "grad_norm": 0.12811633944511414, "learning_rate": 9.51888548541577e-05, "loss": 0.0361, "step": 29400 }, { "epoch": 0.5877063266855841, "grad_norm": 0.1111421138048172, "learning_rate": 9.518443082151421e-05, "loss": 0.0549, "step": 29410 }, { "epoch": 0.5879061588265857, "grad_norm": 0.1515726000070572, "learning_rate": 9.518000485867483e-05, "loss": 0.0364, "step": 29420 }, { "epoch": 0.5881059909675872, "grad_norm": 0.14810512959957123, "learning_rate": 9.517557696582865e-05, "loss": 0.0451, "step": 29430 }, { "epoch": 0.5883058231085888, "grad_norm": 0.17415276169776917, "learning_rate": 9.517114714316481e-05, "loss": 0.0505, "step": 29440 }, { "epoch": 0.5885056552495903, "grad_norm": 0.15404237806797028, "learning_rate": 9.516671539087256e-05, "loss": 0.0359, "step": 29450 }, { "epoch": 0.5887054873905919, "grad_norm": 0.14549694955348969, "learning_rate": 9.516228170914118e-05, "loss": 0.048, "step": 29460 }, { "epoch": 0.5889053195315934, "grad_norm": 0.1591111570596695, "learning_rate": 9.51578460981601e-05, "loss": 0.0484, "step": 29470 }, { "epoch": 0.589105151672595, "grad_norm": 0.15890224277973175, "learning_rate": 9.51534085581188e-05, "loss": 0.0521, "step": 29480 }, { "epoch": 0.5893049838135965, "grad_norm": 0.13632282614707947, "learning_rate": 9.514896908920684e-05, "loss": 0.0366, "step": 29490 }, { "epoch": 0.5895048159545981, "grad_norm": 0.1412135660648346, "learning_rate": 9.514452769161384e-05, "loss": 0.0508, "step": 29500 }, { "epoch": 0.5897046480955997, "grad_norm": 0.12775826454162598, "learning_rate": 9.514008436552956e-05, "loss": 0.0362, "step": 29510 }, { "epoch": 0.5899044802366012, "grad_norm": 0.10878147929906845, "learning_rate": 9.51356391111438e-05, "loss": 0.0403, "step": 29520 }, { "epoch": 0.5901043123776029, "grad_norm": 0.13587719202041626, "learning_rate": 9.513119192864646e-05, "loss": 0.0517, "step": 29530 }, { "epoch": 0.5903041445186044, "grad_norm": 0.15260571241378784, "learning_rate": 9.512674281822751e-05, "loss": 0.0491, "step": 29540 }, { "epoch": 0.590503976659606, "grad_norm": 0.15016500651836395, "learning_rate": 9.512229178007702e-05, "loss": 0.0458, "step": 29550 }, { "epoch": 0.5907038088006075, "grad_norm": 0.1318071484565735, "learning_rate": 9.511783881438509e-05, "loss": 0.0362, "step": 29560 }, { "epoch": 0.5909036409416091, "grad_norm": 0.14473974704742432, "learning_rate": 9.511338392134197e-05, "loss": 0.0438, "step": 29570 }, { "epoch": 0.5911034730826106, "grad_norm": 0.12916536629199982, "learning_rate": 9.510892710113799e-05, "loss": 0.0415, "step": 29580 }, { "epoch": 0.5913033052236122, "grad_norm": 0.14096011221408844, "learning_rate": 9.51044683539635e-05, "loss": 0.039, "step": 29590 }, { "epoch": 0.5915031373646137, "grad_norm": 0.13099314272403717, "learning_rate": 9.510000768000898e-05, "loss": 0.0283, "step": 29600 }, { "epoch": 0.5917029695056153, "grad_norm": 0.1653594821691513, "learning_rate": 9.509554507946496e-05, "loss": 0.0396, "step": 29610 }, { "epoch": 0.5919028016466168, "grad_norm": 0.15219295024871826, "learning_rate": 9.509108055252213e-05, "loss": 0.0387, "step": 29620 }, { "epoch": 0.5921026337876184, "grad_norm": 0.15409061312675476, "learning_rate": 9.508661409937116e-05, "loss": 0.0401, "step": 29630 }, { "epoch": 0.59230246592862, "grad_norm": 0.15343916416168213, "learning_rate": 9.508214572020288e-05, "loss": 0.0393, "step": 29640 }, { "epoch": 0.5925022980696215, "grad_norm": 0.15351086854934692, "learning_rate": 9.507767541520814e-05, "loss": 0.0358, "step": 29650 }, { "epoch": 0.5927021302106231, "grad_norm": 0.13979457318782806, "learning_rate": 9.50732031845779e-05, "loss": 0.0507, "step": 29660 }, { "epoch": 0.5929019623516246, "grad_norm": 0.14774282276630402, "learning_rate": 9.506872902850325e-05, "loss": 0.0379, "step": 29670 }, { "epoch": 0.5931017944926262, "grad_norm": 0.1292254775762558, "learning_rate": 9.506425294717527e-05, "loss": 0.0404, "step": 29680 }, { "epoch": 0.5933016266336277, "grad_norm": 0.13681969046592712, "learning_rate": 9.505977494078519e-05, "loss": 0.0426, "step": 29690 }, { "epoch": 0.5935014587746293, "grad_norm": 0.15588617324829102, "learning_rate": 9.505529500952433e-05, "loss": 0.0396, "step": 29700 }, { "epoch": 0.5937012909156308, "grad_norm": 0.14905162155628204, "learning_rate": 9.505081315358401e-05, "loss": 0.0405, "step": 29710 }, { "epoch": 0.5939011230566325, "grad_norm": 0.15322093665599823, "learning_rate": 9.504632937315571e-05, "loss": 0.0409, "step": 29720 }, { "epoch": 0.594100955197634, "grad_norm": 0.11513492465019226, "learning_rate": 9.504184366843099e-05, "loss": 0.0416, "step": 29730 }, { "epoch": 0.5943007873386356, "grad_norm": 0.20607222616672516, "learning_rate": 9.503735603960144e-05, "loss": 0.0411, "step": 29740 }, { "epoch": 0.5945006194796371, "grad_norm": 0.1591000109910965, "learning_rate": 9.503286648685878e-05, "loss": 0.0441, "step": 29750 }, { "epoch": 0.5947004516206387, "grad_norm": 0.17748618125915527, "learning_rate": 9.502837501039479e-05, "loss": 0.0376, "step": 29760 }, { "epoch": 0.5949002837616402, "grad_norm": 0.11716470867395401, "learning_rate": 9.502388161040134e-05, "loss": 0.0379, "step": 29770 }, { "epoch": 0.5951001159026418, "grad_norm": 0.1378464698791504, "learning_rate": 9.501938628707038e-05, "loss": 0.0407, "step": 29780 }, { "epoch": 0.5952999480436434, "grad_norm": 0.13655675947666168, "learning_rate": 9.501488904059395e-05, "loss": 0.0457, "step": 29790 }, { "epoch": 0.5954997801846449, "grad_norm": 0.10633448511362076, "learning_rate": 9.501038987116415e-05, "loss": 0.0319, "step": 29800 }, { "epoch": 0.5956996123256465, "grad_norm": 0.15759676694869995, "learning_rate": 9.500588877897318e-05, "loss": 0.0397, "step": 29810 }, { "epoch": 0.595899444466648, "grad_norm": 0.1301649510860443, "learning_rate": 9.500138576421332e-05, "loss": 0.0346, "step": 29820 }, { "epoch": 0.5960992766076496, "grad_norm": 0.1464889943599701, "learning_rate": 9.499688082707692e-05, "loss": 0.0509, "step": 29830 }, { "epoch": 0.5962991087486511, "grad_norm": 0.1604321300983429, "learning_rate": 9.499237396775645e-05, "loss": 0.054, "step": 29840 }, { "epoch": 0.5964989408896527, "grad_norm": 0.17319542169570923, "learning_rate": 9.49878651864444e-05, "loss": 0.0402, "step": 29850 }, { "epoch": 0.5966987730306542, "grad_norm": 0.19040878117084503, "learning_rate": 9.498335448333342e-05, "loss": 0.0361, "step": 29860 }, { "epoch": 0.5968986051716558, "grad_norm": 0.16226190328598022, "learning_rate": 9.497884185861614e-05, "loss": 0.0469, "step": 29870 }, { "epoch": 0.5970984373126573, "grad_norm": 0.15569671988487244, "learning_rate": 9.497432731248539e-05, "loss": 0.0428, "step": 29880 }, { "epoch": 0.5972982694536589, "grad_norm": 0.16070587933063507, "learning_rate": 9.496981084513399e-05, "loss": 0.0425, "step": 29890 }, { "epoch": 0.5974981015946604, "grad_norm": 0.14825554192066193, "learning_rate": 9.496529245675487e-05, "loss": 0.0374, "step": 29900 }, { "epoch": 0.597697933735662, "grad_norm": 0.1376456320285797, "learning_rate": 9.496077214754108e-05, "loss": 0.0363, "step": 29910 }, { "epoch": 0.5978977658766637, "grad_norm": 0.12702949345111847, "learning_rate": 9.495624991768568e-05, "loss": 0.0359, "step": 29920 }, { "epoch": 0.5980975980176652, "grad_norm": 0.12839359045028687, "learning_rate": 9.49517257673819e-05, "loss": 0.0351, "step": 29930 }, { "epoch": 0.5982974301586668, "grad_norm": 0.11872268468141556, "learning_rate": 9.494719969682296e-05, "loss": 0.0355, "step": 29940 }, { "epoch": 0.5984972622996683, "grad_norm": 0.12119005620479584, "learning_rate": 9.494267170620222e-05, "loss": 0.0413, "step": 29950 }, { "epoch": 0.5986970944406699, "grad_norm": 0.14724285900592804, "learning_rate": 9.49381417957131e-05, "loss": 0.033, "step": 29960 }, { "epoch": 0.5988969265816714, "grad_norm": 0.15607278048992157, "learning_rate": 9.493360996554912e-05, "loss": 0.0445, "step": 29970 }, { "epoch": 0.599096758722673, "grad_norm": 0.13882039487361908, "learning_rate": 9.492907621590385e-05, "loss": 0.0459, "step": 29980 }, { "epoch": 0.5992965908636745, "grad_norm": 0.12324797362089157, "learning_rate": 9.492454054697101e-05, "loss": 0.0354, "step": 29990 }, { "epoch": 0.5994964230046761, "grad_norm": 0.17552071809768677, "learning_rate": 9.492000295894432e-05, "loss": 0.0419, "step": 30000 }, { "epoch": 0.5996962551456776, "grad_norm": 0.13764460384845734, "learning_rate": 9.491546345201762e-05, "loss": 0.0372, "step": 30010 }, { "epoch": 0.5998960872866792, "grad_norm": 0.14069785177707672, "learning_rate": 9.491092202638483e-05, "loss": 0.0376, "step": 30020 }, { "epoch": 0.6000959194276807, "grad_norm": 0.14630882441997528, "learning_rate": 9.490637868223996e-05, "loss": 0.0368, "step": 30030 }, { "epoch": 0.6002957515686823, "grad_norm": 0.12643195688724518, "learning_rate": 9.49018334197771e-05, "loss": 0.0375, "step": 30040 }, { "epoch": 0.6004955837096838, "grad_norm": 0.1470547914505005, "learning_rate": 9.489728623919039e-05, "loss": 0.0404, "step": 30050 }, { "epoch": 0.6006954158506854, "grad_norm": 0.13335981965065002, "learning_rate": 9.489273714067409e-05, "loss": 0.0436, "step": 30060 }, { "epoch": 0.600895247991687, "grad_norm": 0.13868293166160583, "learning_rate": 9.488818612442252e-05, "loss": 0.0406, "step": 30070 }, { "epoch": 0.6010950801326885, "grad_norm": 0.14432111382484436, "learning_rate": 9.488363319063012e-05, "loss": 0.0411, "step": 30080 }, { "epoch": 0.6012949122736901, "grad_norm": 0.11900264769792557, "learning_rate": 9.487907833949134e-05, "loss": 0.0421, "step": 30090 }, { "epoch": 0.6014947444146916, "grad_norm": 0.13363692164421082, "learning_rate": 9.487452157120081e-05, "loss": 0.0388, "step": 30100 }, { "epoch": 0.6016945765556933, "grad_norm": 0.12652622163295746, "learning_rate": 9.486996288595314e-05, "loss": 0.0352, "step": 30110 }, { "epoch": 0.6018944086966947, "grad_norm": 0.1261025369167328, "learning_rate": 9.486540228394309e-05, "loss": 0.0403, "step": 30120 }, { "epoch": 0.6020942408376964, "grad_norm": 0.2163049280643463, "learning_rate": 9.486083976536546e-05, "loss": 0.0444, "step": 30130 }, { "epoch": 0.6022940729786979, "grad_norm": 0.12640023231506348, "learning_rate": 9.485627533041519e-05, "loss": 0.0406, "step": 30140 }, { "epoch": 0.6024939051196995, "grad_norm": 0.1609897017478943, "learning_rate": 9.485170897928723e-05, "loss": 0.0395, "step": 30150 }, { "epoch": 0.602693737260701, "grad_norm": 0.1442471295595169, "learning_rate": 9.484714071217665e-05, "loss": 0.0359, "step": 30160 }, { "epoch": 0.6028935694017026, "grad_norm": 0.15919700264930725, "learning_rate": 9.48425705292786e-05, "loss": 0.0392, "step": 30170 }, { "epoch": 0.6030934015427041, "grad_norm": 0.18056067824363708, "learning_rate": 9.483799843078835e-05, "loss": 0.037, "step": 30180 }, { "epoch": 0.6032932336837057, "grad_norm": 0.1632179319858551, "learning_rate": 9.483342441690115e-05, "loss": 0.0435, "step": 30190 }, { "epoch": 0.6034930658247073, "grad_norm": 0.13337922096252441, "learning_rate": 9.482884848781242e-05, "loss": 0.0373, "step": 30200 }, { "epoch": 0.6036928979657088, "grad_norm": 0.1337951421737671, "learning_rate": 9.482427064371764e-05, "loss": 0.0394, "step": 30210 }, { "epoch": 0.6038927301067104, "grad_norm": 0.15756702423095703, "learning_rate": 9.481969088481237e-05, "loss": 0.0515, "step": 30220 }, { "epoch": 0.6040925622477119, "grad_norm": 0.14680524170398712, "learning_rate": 9.481510921129224e-05, "loss": 0.0488, "step": 30230 }, { "epoch": 0.6042923943887135, "grad_norm": 0.15835201740264893, "learning_rate": 9.481052562335297e-05, "loss": 0.0379, "step": 30240 }, { "epoch": 0.604492226529715, "grad_norm": 0.13785448670387268, "learning_rate": 9.480594012119035e-05, "loss": 0.0435, "step": 30250 }, { "epoch": 0.6046920586707166, "grad_norm": 0.15107905864715576, "learning_rate": 9.48013527050003e-05, "loss": 0.0418, "step": 30260 }, { "epoch": 0.6048918908117181, "grad_norm": 0.14834149181842804, "learning_rate": 9.479676337497875e-05, "loss": 0.0372, "step": 30270 }, { "epoch": 0.6050917229527197, "grad_norm": 0.14415395259857178, "learning_rate": 9.479217213132178e-05, "loss": 0.0334, "step": 30280 }, { "epoch": 0.6052915550937212, "grad_norm": 0.1057390496134758, "learning_rate": 9.47875789742255e-05, "loss": 0.0417, "step": 30290 }, { "epoch": 0.6054913872347228, "grad_norm": 0.1300869584083557, "learning_rate": 9.478298390388611e-05, "loss": 0.0369, "step": 30300 }, { "epoch": 0.6056912193757243, "grad_norm": 0.12715746462345123, "learning_rate": 9.477838692049992e-05, "loss": 0.0384, "step": 30310 }, { "epoch": 0.605891051516726, "grad_norm": 0.1561007797718048, "learning_rate": 9.477378802426331e-05, "loss": 0.0372, "step": 30320 }, { "epoch": 0.6060908836577276, "grad_norm": 0.1444145292043686, "learning_rate": 9.476918721537271e-05, "loss": 0.0312, "step": 30330 }, { "epoch": 0.6062907157987291, "grad_norm": 0.12286559492349625, "learning_rate": 9.47645844940247e-05, "loss": 0.0463, "step": 30340 }, { "epoch": 0.6064905479397307, "grad_norm": 0.10945624858140945, "learning_rate": 9.475997986041586e-05, "loss": 0.0346, "step": 30350 }, { "epoch": 0.6066903800807322, "grad_norm": 0.13958603143692017, "learning_rate": 9.475537331474291e-05, "loss": 0.0405, "step": 30360 }, { "epoch": 0.6068902122217338, "grad_norm": 0.1568276286125183, "learning_rate": 9.475076485720262e-05, "loss": 0.0383, "step": 30370 }, { "epoch": 0.6070900443627353, "grad_norm": 0.15488430857658386, "learning_rate": 9.474615448799189e-05, "loss": 0.04, "step": 30380 }, { "epoch": 0.6072898765037369, "grad_norm": 0.13935135304927826, "learning_rate": 9.474154220730762e-05, "loss": 0.0465, "step": 30390 }, { "epoch": 0.6074897086447384, "grad_norm": 0.15927928686141968, "learning_rate": 9.473692801534687e-05, "loss": 0.0419, "step": 30400 }, { "epoch": 0.60768954078574, "grad_norm": 0.11714455485343933, "learning_rate": 9.473231191230673e-05, "loss": 0.0383, "step": 30410 }, { "epoch": 0.6078893729267415, "grad_norm": 0.11382682621479034, "learning_rate": 9.472769389838442e-05, "loss": 0.0381, "step": 30420 }, { "epoch": 0.6080892050677431, "grad_norm": 0.19706954061985016, "learning_rate": 9.472307397377717e-05, "loss": 0.0407, "step": 30430 }, { "epoch": 0.6082890372087446, "grad_norm": 0.1224679946899414, "learning_rate": 9.471845213868238e-05, "loss": 0.0361, "step": 30440 }, { "epoch": 0.6084888693497462, "grad_norm": 0.16817204654216766, "learning_rate": 9.471382839329745e-05, "loss": 0.0376, "step": 30450 }, { "epoch": 0.6086887014907477, "grad_norm": 0.13393090665340424, "learning_rate": 9.470920273781993e-05, "loss": 0.038, "step": 30460 }, { "epoch": 0.6088885336317493, "grad_norm": 0.15484382212162018, "learning_rate": 9.470457517244738e-05, "loss": 0.0445, "step": 30470 }, { "epoch": 0.6090883657727509, "grad_norm": 0.1318926364183426, "learning_rate": 9.469994569737753e-05, "loss": 0.0438, "step": 30480 }, { "epoch": 0.6092881979137524, "grad_norm": 0.10186930000782013, "learning_rate": 9.46953143128081e-05, "loss": 0.0478, "step": 30490 }, { "epoch": 0.609488030054754, "grad_norm": 0.1355738341808319, "learning_rate": 9.469068101893694e-05, "loss": 0.0394, "step": 30500 }, { "epoch": 0.6096878621957555, "grad_norm": 0.13356798887252808, "learning_rate": 9.468604581596201e-05, "loss": 0.0355, "step": 30510 }, { "epoch": 0.6098876943367572, "grad_norm": 0.1199152022600174, "learning_rate": 9.468140870408126e-05, "loss": 0.0476, "step": 30520 }, { "epoch": 0.6100875264777587, "grad_norm": 0.14781291782855988, "learning_rate": 9.467676968349284e-05, "loss": 0.0482, "step": 30530 }, { "epoch": 0.6102873586187603, "grad_norm": 0.1742587387561798, "learning_rate": 9.46721287543949e-05, "loss": 0.0393, "step": 30540 }, { "epoch": 0.6104871907597618, "grad_norm": 0.13433560729026794, "learning_rate": 9.466748591698566e-05, "loss": 0.0402, "step": 30550 }, { "epoch": 0.6106870229007634, "grad_norm": 0.11347655951976776, "learning_rate": 9.46628411714635e-05, "loss": 0.0377, "step": 30560 }, { "epoch": 0.6108868550417649, "grad_norm": 0.1948913186788559, "learning_rate": 9.46581945180268e-05, "loss": 0.042, "step": 30570 }, { "epoch": 0.6110866871827665, "grad_norm": 0.14721813797950745, "learning_rate": 9.465354595687407e-05, "loss": 0.0405, "step": 30580 }, { "epoch": 0.611286519323768, "grad_norm": 0.15177398920059204, "learning_rate": 9.46488954882039e-05, "loss": 0.0398, "step": 30590 }, { "epoch": 0.6114863514647696, "grad_norm": 0.14655379951000214, "learning_rate": 9.464424311221493e-05, "loss": 0.0401, "step": 30600 }, { "epoch": 0.6116861836057712, "grad_norm": 0.1649317592382431, "learning_rate": 9.46395888291059e-05, "loss": 0.0422, "step": 30610 }, { "epoch": 0.6118860157467727, "grad_norm": 0.14750947058200836, "learning_rate": 9.463493263907564e-05, "loss": 0.049, "step": 30620 }, { "epoch": 0.6120858478877743, "grad_norm": 0.16509248316287994, "learning_rate": 9.463027454232307e-05, "loss": 0.048, "step": 30630 }, { "epoch": 0.6122856800287758, "grad_norm": 0.13187769055366516, "learning_rate": 9.462561453904714e-05, "loss": 0.0511, "step": 30640 }, { "epoch": 0.6124855121697774, "grad_norm": 0.16154591739177704, "learning_rate": 9.462095262944696e-05, "loss": 0.0492, "step": 30650 }, { "epoch": 0.6126853443107789, "grad_norm": 0.1432584971189499, "learning_rate": 9.461628881372164e-05, "loss": 0.0422, "step": 30660 }, { "epoch": 0.6128851764517805, "grad_norm": 0.16494740545749664, "learning_rate": 9.461162309207043e-05, "loss": 0.0431, "step": 30670 }, { "epoch": 0.613085008592782, "grad_norm": 0.13532428443431854, "learning_rate": 9.460695546469264e-05, "loss": 0.0427, "step": 30680 }, { "epoch": 0.6132848407337836, "grad_norm": 0.1402513086795807, "learning_rate": 9.460228593178766e-05, "loss": 0.0402, "step": 30690 }, { "epoch": 0.6134846728747851, "grad_norm": 0.17736078798770905, "learning_rate": 9.459761449355496e-05, "loss": 0.0383, "step": 30700 }, { "epoch": 0.6136845050157868, "grad_norm": 0.14535373449325562, "learning_rate": 9.459294115019408e-05, "loss": 0.0407, "step": 30710 }, { "epoch": 0.6138843371567883, "grad_norm": 0.16350504755973816, "learning_rate": 9.458826590190468e-05, "loss": 0.0372, "step": 30720 }, { "epoch": 0.6140841692977899, "grad_norm": 0.18206100165843964, "learning_rate": 9.458358874888649e-05, "loss": 0.0338, "step": 30730 }, { "epoch": 0.6142840014387914, "grad_norm": 0.18806473910808563, "learning_rate": 9.457890969133927e-05, "loss": 0.0394, "step": 30740 }, { "epoch": 0.614483833579793, "grad_norm": 0.12127834558486938, "learning_rate": 9.457422872946293e-05, "loss": 0.0416, "step": 30750 }, { "epoch": 0.6146836657207946, "grad_norm": 0.2809220552444458, "learning_rate": 9.456954586345743e-05, "loss": 0.0425, "step": 30760 }, { "epoch": 0.6148834978617961, "grad_norm": 0.11363490670919418, "learning_rate": 9.45648610935228e-05, "loss": 0.0401, "step": 30770 }, { "epoch": 0.6150833300027977, "grad_norm": 0.1367308497428894, "learning_rate": 9.456017441985919e-05, "loss": 0.0558, "step": 30780 }, { "epoch": 0.6152831621437992, "grad_norm": 0.15900073945522308, "learning_rate": 9.455548584266676e-05, "loss": 0.0412, "step": 30790 }, { "epoch": 0.6154829942848008, "grad_norm": 0.15346117317676544, "learning_rate": 9.455079536214584e-05, "loss": 0.0505, "step": 30800 }, { "epoch": 0.6156828264258023, "grad_norm": 0.1550903171300888, "learning_rate": 9.454610297849678e-05, "loss": 0.0407, "step": 30810 }, { "epoch": 0.6158826585668039, "grad_norm": 0.1365959644317627, "learning_rate": 9.454140869192006e-05, "loss": 0.0339, "step": 30820 }, { "epoch": 0.6160824907078054, "grad_norm": 0.1420353651046753, "learning_rate": 9.453671250261616e-05, "loss": 0.0538, "step": 30830 }, { "epoch": 0.616282322848807, "grad_norm": 0.168405219912529, "learning_rate": 9.453201441078572e-05, "loss": 0.0388, "step": 30840 }, { "epoch": 0.6164821549898085, "grad_norm": 0.12871168553829193, "learning_rate": 9.452731441662944e-05, "loss": 0.0334, "step": 30850 }, { "epoch": 0.6166819871308101, "grad_norm": 0.12372247129678726, "learning_rate": 9.452261252034807e-05, "loss": 0.0378, "step": 30860 }, { "epoch": 0.6168818192718116, "grad_norm": 0.1596732884645462, "learning_rate": 9.45179087221425e-05, "loss": 0.0427, "step": 30870 }, { "epoch": 0.6170816514128132, "grad_norm": 0.13358472287654877, "learning_rate": 9.451320302221365e-05, "loss": 0.0357, "step": 30880 }, { "epoch": 0.6172814835538148, "grad_norm": 0.13464204967021942, "learning_rate": 9.450849542076254e-05, "loss": 0.0479, "step": 30890 }, { "epoch": 0.6174813156948163, "grad_norm": 0.16153931617736816, "learning_rate": 9.450378591799027e-05, "loss": 0.0344, "step": 30900 }, { "epoch": 0.617681147835818, "grad_norm": 0.1327829211950302, "learning_rate": 9.449907451409802e-05, "loss": 0.0398, "step": 30910 }, { "epoch": 0.6178809799768195, "grad_norm": 0.17227137088775635, "learning_rate": 9.449436120928706e-05, "loss": 0.0408, "step": 30920 }, { "epoch": 0.6180808121178211, "grad_norm": 0.14006327092647552, "learning_rate": 9.448964600375872e-05, "loss": 0.0415, "step": 30930 }, { "epoch": 0.6182806442588226, "grad_norm": 0.10397368669509888, "learning_rate": 9.448492889771443e-05, "loss": 0.0433, "step": 30940 }, { "epoch": 0.6184804763998242, "grad_norm": 0.14395511150360107, "learning_rate": 9.448020989135571e-05, "loss": 0.0386, "step": 30950 }, { "epoch": 0.6186803085408257, "grad_norm": 0.10037161409854889, "learning_rate": 9.447548898488412e-05, "loss": 0.0346, "step": 30960 }, { "epoch": 0.6188801406818273, "grad_norm": 0.21495340764522552, "learning_rate": 9.447076617850136e-05, "loss": 0.046, "step": 30970 }, { "epoch": 0.6190799728228288, "grad_norm": 0.16946706175804138, "learning_rate": 9.446604147240917e-05, "loss": 0.0459, "step": 30980 }, { "epoch": 0.6192798049638304, "grad_norm": 0.18677237629890442, "learning_rate": 9.446131486680936e-05, "loss": 0.0364, "step": 30990 }, { "epoch": 0.6194796371048319, "grad_norm": 0.1492013782262802, "learning_rate": 9.445658636190385e-05, "loss": 0.0386, "step": 31000 }, { "epoch": 0.6196794692458335, "grad_norm": 0.16257943212985992, "learning_rate": 9.445185595789466e-05, "loss": 0.0394, "step": 31010 }, { "epoch": 0.6198793013868351, "grad_norm": 0.19158847630023956, "learning_rate": 9.444712365498382e-05, "loss": 0.0393, "step": 31020 }, { "epoch": 0.6200791335278366, "grad_norm": 0.17964842915534973, "learning_rate": 9.444238945337354e-05, "loss": 0.0422, "step": 31030 }, { "epoch": 0.6202789656688382, "grad_norm": 0.1381094604730606, "learning_rate": 9.4437653353266e-05, "loss": 0.0395, "step": 31040 }, { "epoch": 0.6204787978098397, "grad_norm": 0.17516633868217468, "learning_rate": 9.443291535486355e-05, "loss": 0.0516, "step": 31050 }, { "epoch": 0.6206786299508413, "grad_norm": 0.19401641190052032, "learning_rate": 9.442817545836859e-05, "loss": 0.0429, "step": 31060 }, { "epoch": 0.6208784620918428, "grad_norm": 0.17700183391571045, "learning_rate": 9.442343366398359e-05, "loss": 0.0424, "step": 31070 }, { "epoch": 0.6210782942328444, "grad_norm": 0.151814267039299, "learning_rate": 9.441868997191111e-05, "loss": 0.0496, "step": 31080 }, { "epoch": 0.6212781263738459, "grad_norm": 0.20049692690372467, "learning_rate": 9.441394438235379e-05, "loss": 0.0408, "step": 31090 }, { "epoch": 0.6214779585148476, "grad_norm": 0.13163210451602936, "learning_rate": 9.440919689551436e-05, "loss": 0.036, "step": 31100 }, { "epoch": 0.621677790655849, "grad_norm": 0.15266340970993042, "learning_rate": 9.440444751159562e-05, "loss": 0.0386, "step": 31110 }, { "epoch": 0.6218776227968507, "grad_norm": 0.1252821683883667, "learning_rate": 9.439969623080047e-05, "loss": 0.0348, "step": 31120 }, { "epoch": 0.6220774549378522, "grad_norm": 0.15032479166984558, "learning_rate": 9.439494305333185e-05, "loss": 0.0371, "step": 31130 }, { "epoch": 0.6222772870788538, "grad_norm": 0.14854781329631805, "learning_rate": 9.43901879793928e-05, "loss": 0.0353, "step": 31140 }, { "epoch": 0.6224771192198553, "grad_norm": 0.1447819620370865, "learning_rate": 9.43854310091865e-05, "loss": 0.041, "step": 31150 }, { "epoch": 0.6226769513608569, "grad_norm": 0.13657133281230927, "learning_rate": 9.43806721429161e-05, "loss": 0.0389, "step": 31160 }, { "epoch": 0.6228767835018585, "grad_norm": 0.15732643008232117, "learning_rate": 9.437591138078494e-05, "loss": 0.0434, "step": 31170 }, { "epoch": 0.62307661564286, "grad_norm": 0.13177548348903656, "learning_rate": 9.437114872299636e-05, "loss": 0.0416, "step": 31180 }, { "epoch": 0.6232764477838616, "grad_norm": 0.16896800696849823, "learning_rate": 9.436638416975381e-05, "loss": 0.0411, "step": 31190 }, { "epoch": 0.6234762799248631, "grad_norm": 0.12599003314971924, "learning_rate": 9.436161772126082e-05, "loss": 0.0338, "step": 31200 }, { "epoch": 0.6236761120658647, "grad_norm": 0.12089799344539642, "learning_rate": 9.435684937772105e-05, "loss": 0.0348, "step": 31210 }, { "epoch": 0.6238759442068662, "grad_norm": 0.12308301031589508, "learning_rate": 9.435207913933813e-05, "loss": 0.0417, "step": 31220 }, { "epoch": 0.6240757763478678, "grad_norm": 0.14244607090950012, "learning_rate": 9.434730700631589e-05, "loss": 0.0416, "step": 31230 }, { "epoch": 0.6242756084888693, "grad_norm": 0.11132621765136719, "learning_rate": 9.434253297885814e-05, "loss": 0.0407, "step": 31240 }, { "epoch": 0.6244754406298709, "grad_norm": 0.1836550533771515, "learning_rate": 9.433775705716885e-05, "loss": 0.0291, "step": 31250 }, { "epoch": 0.6246752727708724, "grad_norm": 0.13740187883377075, "learning_rate": 9.433297924145204e-05, "loss": 0.0367, "step": 31260 }, { "epoch": 0.624875104911874, "grad_norm": 0.14502671360969543, "learning_rate": 9.432819953191178e-05, "loss": 0.0378, "step": 31270 }, { "epoch": 0.6250749370528755, "grad_norm": 0.16566286981105804, "learning_rate": 9.432341792875229e-05, "loss": 0.0415, "step": 31280 }, { "epoch": 0.6252747691938771, "grad_norm": 0.14763198792934418, "learning_rate": 9.43186344321778e-05, "loss": 0.0474, "step": 31290 }, { "epoch": 0.6254746013348788, "grad_norm": 0.14813192188739777, "learning_rate": 9.431384904239267e-05, "loss": 0.0387, "step": 31300 }, { "epoch": 0.6256744334758803, "grad_norm": 0.13666562736034393, "learning_rate": 9.43090617596013e-05, "loss": 0.0364, "step": 31310 }, { "epoch": 0.6258742656168819, "grad_norm": 0.11680545657873154, "learning_rate": 9.430427258400822e-05, "loss": 0.0411, "step": 31320 }, { "epoch": 0.6260740977578834, "grad_norm": 0.12750118970870972, "learning_rate": 9.429948151581801e-05, "loss": 0.0382, "step": 31330 }, { "epoch": 0.626273929898885, "grad_norm": 0.17783884704113007, "learning_rate": 9.429468855523532e-05, "loss": 0.0406, "step": 31340 }, { "epoch": 0.6264737620398865, "grad_norm": 0.11962484568357468, "learning_rate": 9.428989370246492e-05, "loss": 0.0375, "step": 31350 }, { "epoch": 0.6266735941808881, "grad_norm": 0.15157203376293182, "learning_rate": 9.428509695771162e-05, "loss": 0.0385, "step": 31360 }, { "epoch": 0.6268734263218896, "grad_norm": 0.1615740805864334, "learning_rate": 9.428029832118033e-05, "loss": 0.0381, "step": 31370 }, { "epoch": 0.6270732584628912, "grad_norm": 0.1455371379852295, "learning_rate": 9.427549779307603e-05, "loss": 0.0446, "step": 31380 }, { "epoch": 0.6272730906038927, "grad_norm": 0.1421583741903305, "learning_rate": 9.427069537360381e-05, "loss": 0.0389, "step": 31390 }, { "epoch": 0.6274729227448943, "grad_norm": 0.14663781225681305, "learning_rate": 9.426589106296882e-05, "loss": 0.0375, "step": 31400 }, { "epoch": 0.6276727548858958, "grad_norm": 0.12289492040872574, "learning_rate": 9.426108486137628e-05, "loss": 0.0483, "step": 31410 }, { "epoch": 0.6278725870268974, "grad_norm": 0.12963378429412842, "learning_rate": 9.42562767690315e-05, "loss": 0.0357, "step": 31420 }, { "epoch": 0.6280724191678989, "grad_norm": 0.14628712832927704, "learning_rate": 9.425146678613987e-05, "loss": 0.0471, "step": 31430 }, { "epoch": 0.6282722513089005, "grad_norm": 0.15809687972068787, "learning_rate": 9.424665491290687e-05, "loss": 0.0396, "step": 31440 }, { "epoch": 0.6284720834499021, "grad_norm": 0.1698402315378189, "learning_rate": 9.424184114953806e-05, "loss": 0.0372, "step": 31450 }, { "epoch": 0.6286719155909036, "grad_norm": 0.17598684132099152, "learning_rate": 9.42370254962391e-05, "loss": 0.0368, "step": 31460 }, { "epoch": 0.6288717477319052, "grad_norm": 0.1286027729511261, "learning_rate": 9.423220795321564e-05, "loss": 0.0417, "step": 31470 }, { "epoch": 0.6290715798729067, "grad_norm": 0.12469395250082016, "learning_rate": 9.422738852067354e-05, "loss": 0.0435, "step": 31480 }, { "epoch": 0.6292714120139083, "grad_norm": 0.1381230503320694, "learning_rate": 9.422256719881864e-05, "loss": 0.0393, "step": 31490 }, { "epoch": 0.6294712441549098, "grad_norm": 0.15800394117832184, "learning_rate": 9.42177439878569e-05, "loss": 0.0384, "step": 31500 }, { "epoch": 0.6296710762959115, "grad_norm": 0.14946193993091583, "learning_rate": 9.421291888799437e-05, "loss": 0.0417, "step": 31510 }, { "epoch": 0.629870908436913, "grad_norm": 0.12881651520729065, "learning_rate": 9.420809189943717e-05, "loss": 0.0423, "step": 31520 }, { "epoch": 0.6300707405779146, "grad_norm": 0.13421620428562164, "learning_rate": 9.42032630223915e-05, "loss": 0.0441, "step": 31530 }, { "epoch": 0.6302705727189161, "grad_norm": 0.1446388214826584, "learning_rate": 9.419843225706363e-05, "loss": 0.0451, "step": 31540 }, { "epoch": 0.6304704048599177, "grad_norm": 0.13064290583133698, "learning_rate": 9.419359960365995e-05, "loss": 0.0435, "step": 31550 }, { "epoch": 0.6306702370009192, "grad_norm": 0.1820686310529709, "learning_rate": 9.418876506238686e-05, "loss": 0.046, "step": 31560 }, { "epoch": 0.6308700691419208, "grad_norm": 0.10675965994596481, "learning_rate": 9.41839286334509e-05, "loss": 0.0356, "step": 31570 }, { "epoch": 0.6310699012829224, "grad_norm": 0.16234412789344788, "learning_rate": 9.417909031705868e-05, "loss": 0.0455, "step": 31580 }, { "epoch": 0.6312697334239239, "grad_norm": 0.1329144388437271, "learning_rate": 9.41742501134169e-05, "loss": 0.0383, "step": 31590 }, { "epoch": 0.6314695655649255, "grad_norm": 0.1329997032880783, "learning_rate": 9.416940802273229e-05, "loss": 0.049, "step": 31600 }, { "epoch": 0.631669397705927, "grad_norm": 0.140178844332695, "learning_rate": 9.416456404521172e-05, "loss": 0.0408, "step": 31610 }, { "epoch": 0.6318692298469286, "grad_norm": 0.12211780250072479, "learning_rate": 9.41597181810621e-05, "loss": 0.0389, "step": 31620 }, { "epoch": 0.6320690619879301, "grad_norm": 0.12219199538230896, "learning_rate": 9.415487043049044e-05, "loss": 0.0522, "step": 31630 }, { "epoch": 0.6322688941289317, "grad_norm": 0.1763693243265152, "learning_rate": 9.415002079370383e-05, "loss": 0.038, "step": 31640 }, { "epoch": 0.6324687262699332, "grad_norm": 0.13143862783908844, "learning_rate": 9.414516927090944e-05, "loss": 0.0431, "step": 31650 }, { "epoch": 0.6326685584109348, "grad_norm": 0.1646997183561325, "learning_rate": 9.414031586231452e-05, "loss": 0.0417, "step": 31660 }, { "epoch": 0.6328683905519363, "grad_norm": 0.15378707647323608, "learning_rate": 9.41354605681264e-05, "loss": 0.0354, "step": 31670 }, { "epoch": 0.6330682226929379, "grad_norm": 0.14202837646007538, "learning_rate": 9.413060338855247e-05, "loss": 0.04, "step": 31680 }, { "epoch": 0.6332680548339394, "grad_norm": 0.17157302796840668, "learning_rate": 9.412574432380025e-05, "loss": 0.0496, "step": 31690 }, { "epoch": 0.633467886974941, "grad_norm": 0.12945814430713654, "learning_rate": 9.412088337407728e-05, "loss": 0.032, "step": 31700 }, { "epoch": 0.6336677191159427, "grad_norm": 0.1512453854084015, "learning_rate": 9.411602053959123e-05, "loss": 0.0412, "step": 31710 }, { "epoch": 0.6338675512569442, "grad_norm": 0.13011421263217926, "learning_rate": 9.411115582054983e-05, "loss": 0.0341, "step": 31720 }, { "epoch": 0.6340673833979458, "grad_norm": 0.1463659703731537, "learning_rate": 9.410628921716088e-05, "loss": 0.0406, "step": 31730 }, { "epoch": 0.6342672155389473, "grad_norm": 0.14625057578086853, "learning_rate": 9.410142072963228e-05, "loss": 0.0356, "step": 31740 }, { "epoch": 0.6344670476799489, "grad_norm": 0.13928453624248505, "learning_rate": 9.409655035817201e-05, "loss": 0.0581, "step": 31750 }, { "epoch": 0.6346668798209504, "grad_norm": 0.11660575121641159, "learning_rate": 9.409167810298811e-05, "loss": 0.0389, "step": 31760 }, { "epoch": 0.634866711961952, "grad_norm": 0.13208334147930145, "learning_rate": 9.408680396428874e-05, "loss": 0.033, "step": 31770 }, { "epoch": 0.6350665441029535, "grad_norm": 0.14978563785552979, "learning_rate": 9.408192794228208e-05, "loss": 0.043, "step": 31780 }, { "epoch": 0.6352663762439551, "grad_norm": 0.12582695484161377, "learning_rate": 9.407705003717644e-05, "loss": 0.0415, "step": 31790 }, { "epoch": 0.6354662083849566, "grad_norm": 0.13540427386760712, "learning_rate": 9.40721702491802e-05, "loss": 0.0428, "step": 31800 }, { "epoch": 0.6356660405259582, "grad_norm": 0.1402401626110077, "learning_rate": 9.40672885785018e-05, "loss": 0.0455, "step": 31810 }, { "epoch": 0.6358658726669597, "grad_norm": 0.12111934274435043, "learning_rate": 9.406240502534979e-05, "loss": 0.0421, "step": 31820 }, { "epoch": 0.6360657048079613, "grad_norm": 0.14900685846805573, "learning_rate": 9.405751958993279e-05, "loss": 0.0393, "step": 31830 }, { "epoch": 0.6362655369489628, "grad_norm": 0.12781886756420135, "learning_rate": 9.405263227245946e-05, "loss": 0.0403, "step": 31840 }, { "epoch": 0.6364653690899644, "grad_norm": 0.17893773317337036, "learning_rate": 9.404774307313864e-05, "loss": 0.0613, "step": 31850 }, { "epoch": 0.636665201230966, "grad_norm": 0.16343368589878082, "learning_rate": 9.404285199217915e-05, "loss": 0.0457, "step": 31860 }, { "epoch": 0.6368650333719675, "grad_norm": 0.10292911529541016, "learning_rate": 9.403795902978992e-05, "loss": 0.0328, "step": 31870 }, { "epoch": 0.6370648655129691, "grad_norm": 0.14386336505413055, "learning_rate": 9.403306418617998e-05, "loss": 0.0431, "step": 31880 }, { "epoch": 0.6372646976539706, "grad_norm": 0.16927090287208557, "learning_rate": 9.402816746155844e-05, "loss": 0.0432, "step": 31890 }, { "epoch": 0.6374645297949723, "grad_norm": 0.1330779492855072, "learning_rate": 9.402326885613446e-05, "loss": 0.0419, "step": 31900 }, { "epoch": 0.6376643619359738, "grad_norm": 0.10972286760807037, "learning_rate": 9.401836837011732e-05, "loss": 0.0469, "step": 31910 }, { "epoch": 0.6378641940769754, "grad_norm": 0.15069250762462616, "learning_rate": 9.401346600371634e-05, "loss": 0.0404, "step": 31920 }, { "epoch": 0.6380640262179769, "grad_norm": 0.1543377935886383, "learning_rate": 9.400856175714094e-05, "loss": 0.0463, "step": 31930 }, { "epoch": 0.6382638583589785, "grad_norm": 0.1676538586616516, "learning_rate": 9.400365563060063e-05, "loss": 0.0381, "step": 31940 }, { "epoch": 0.63846369049998, "grad_norm": 0.19865098595619202, "learning_rate": 9.3998747624305e-05, "loss": 0.0368, "step": 31950 }, { "epoch": 0.6386635226409816, "grad_norm": 0.09902214258909225, "learning_rate": 9.399383773846369e-05, "loss": 0.0462, "step": 31960 }, { "epoch": 0.6388633547819831, "grad_norm": 0.13045448064804077, "learning_rate": 9.398892597328646e-05, "loss": 0.0369, "step": 31970 }, { "epoch": 0.6390631869229847, "grad_norm": 0.15870392322540283, "learning_rate": 9.39840123289831e-05, "loss": 0.0415, "step": 31980 }, { "epoch": 0.6392630190639863, "grad_norm": 0.16064682602882385, "learning_rate": 9.397909680576355e-05, "loss": 0.0385, "step": 31990 }, { "epoch": 0.6394628512049878, "grad_norm": 0.13558271527290344, "learning_rate": 9.397417940383779e-05, "loss": 0.0366, "step": 32000 }, { "epoch": 0.6396626833459894, "grad_norm": 0.14943723380565643, "learning_rate": 9.396926012341586e-05, "loss": 0.0356, "step": 32010 }, { "epoch": 0.6398625154869909, "grad_norm": 0.12884168326854706, "learning_rate": 9.396433896470792e-05, "loss": 0.0391, "step": 32020 }, { "epoch": 0.6400623476279925, "grad_norm": 0.11789079010486603, "learning_rate": 9.395941592792417e-05, "loss": 0.0304, "step": 32030 }, { "epoch": 0.640262179768994, "grad_norm": 0.14186735451221466, "learning_rate": 9.395449101327491e-05, "loss": 0.0382, "step": 32040 }, { "epoch": 0.6404620119099956, "grad_norm": 0.15050075948238373, "learning_rate": 9.394956422097057e-05, "loss": 0.0388, "step": 32050 }, { "epoch": 0.6406618440509971, "grad_norm": 0.1509515643119812, "learning_rate": 9.394463555122158e-05, "loss": 0.0368, "step": 32060 }, { "epoch": 0.6408616761919987, "grad_norm": 0.11595724523067474, "learning_rate": 9.393970500423847e-05, "loss": 0.0391, "step": 32070 }, { "epoch": 0.6410615083330002, "grad_norm": 0.14941959083080292, "learning_rate": 9.393477258023191e-05, "loss": 0.0425, "step": 32080 }, { "epoch": 0.6412613404740019, "grad_norm": 0.16889050602912903, "learning_rate": 9.392983827941256e-05, "loss": 0.042, "step": 32090 }, { "epoch": 0.6414611726150033, "grad_norm": 0.15059910714626312, "learning_rate": 9.392490210199123e-05, "loss": 0.0397, "step": 32100 }, { "epoch": 0.641661004756005, "grad_norm": 0.12548713386058807, "learning_rate": 9.391996404817875e-05, "loss": 0.0387, "step": 32110 }, { "epoch": 0.6418608368970065, "grad_norm": 0.17467671632766724, "learning_rate": 9.391502411818611e-05, "loss": 0.0381, "step": 32120 }, { "epoch": 0.6420606690380081, "grad_norm": 0.1494942456483841, "learning_rate": 9.391008231222431e-05, "loss": 0.0359, "step": 32130 }, { "epoch": 0.6422605011790097, "grad_norm": 0.193076491355896, "learning_rate": 9.390513863050447e-05, "loss": 0.0417, "step": 32140 }, { "epoch": 0.6424603333200112, "grad_norm": 0.17645321786403656, "learning_rate": 9.390019307323776e-05, "loss": 0.052, "step": 32150 }, { "epoch": 0.6426601654610128, "grad_norm": 0.15913331508636475, "learning_rate": 9.389524564063543e-05, "loss": 0.042, "step": 32160 }, { "epoch": 0.6428599976020143, "grad_norm": 0.1709161251783371, "learning_rate": 9.389029633290886e-05, "loss": 0.0457, "step": 32170 }, { "epoch": 0.6430598297430159, "grad_norm": 0.1327994465827942, "learning_rate": 9.388534515026946e-05, "loss": 0.0359, "step": 32180 }, { "epoch": 0.6432596618840174, "grad_norm": 0.16256052255630493, "learning_rate": 9.388039209292874e-05, "loss": 0.0445, "step": 32190 }, { "epoch": 0.643459494025019, "grad_norm": 0.13999584317207336, "learning_rate": 9.387543716109827e-05, "loss": 0.0353, "step": 32200 }, { "epoch": 0.6436593261660205, "grad_norm": 0.14265383780002594, "learning_rate": 9.387048035498973e-05, "loss": 0.0345, "step": 32210 }, { "epoch": 0.6438591583070221, "grad_norm": 0.16267751157283783, "learning_rate": 9.386552167481486e-05, "loss": 0.0389, "step": 32220 }, { "epoch": 0.6440589904480236, "grad_norm": 0.17362593114376068, "learning_rate": 9.386056112078548e-05, "loss": 0.0371, "step": 32230 }, { "epoch": 0.6442588225890252, "grad_norm": 0.14220178127288818, "learning_rate": 9.385559869311352e-05, "loss": 0.0462, "step": 32240 }, { "epoch": 0.6444586547300267, "grad_norm": 0.15386860072612762, "learning_rate": 9.385063439201093e-05, "loss": 0.0364, "step": 32250 }, { "epoch": 0.6446584868710283, "grad_norm": 0.1529797464609146, "learning_rate": 9.384566821768982e-05, "loss": 0.0356, "step": 32260 }, { "epoch": 0.64485831901203, "grad_norm": 0.12339650094509125, "learning_rate": 9.384070017036229e-05, "loss": 0.0443, "step": 32270 }, { "epoch": 0.6450581511530314, "grad_norm": 0.12861381471157074, "learning_rate": 9.383573025024059e-05, "loss": 0.0431, "step": 32280 }, { "epoch": 0.645257983294033, "grad_norm": 0.17089203000068665, "learning_rate": 9.383075845753704e-05, "loss": 0.0401, "step": 32290 }, { "epoch": 0.6454578154350346, "grad_norm": 0.13113363087177277, "learning_rate": 9.382578479246398e-05, "loss": 0.0367, "step": 32300 }, { "epoch": 0.6456576475760362, "grad_norm": 0.13503937423229218, "learning_rate": 9.382080925523393e-05, "loss": 0.0438, "step": 32310 }, { "epoch": 0.6458574797170377, "grad_norm": 0.12049958109855652, "learning_rate": 9.38158318460594e-05, "loss": 0.0421, "step": 32320 }, { "epoch": 0.6460573118580393, "grad_norm": 0.13186705112457275, "learning_rate": 9.381085256515304e-05, "loss": 0.0363, "step": 32330 }, { "epoch": 0.6462571439990408, "grad_norm": 0.1379723846912384, "learning_rate": 9.380587141272751e-05, "loss": 0.0375, "step": 32340 }, { "epoch": 0.6464569761400424, "grad_norm": 0.12737993896007538, "learning_rate": 9.380088838899564e-05, "loss": 0.0491, "step": 32350 }, { "epoch": 0.6466568082810439, "grad_norm": 0.13318127393722534, "learning_rate": 9.379590349417029e-05, "loss": 0.0432, "step": 32360 }, { "epoch": 0.6468566404220455, "grad_norm": 0.13099519908428192, "learning_rate": 9.37909167284644e-05, "loss": 0.0466, "step": 32370 }, { "epoch": 0.647056472563047, "grad_norm": 0.13972334563732147, "learning_rate": 9.378592809209098e-05, "loss": 0.0335, "step": 32380 }, { "epoch": 0.6472563047040486, "grad_norm": 0.1432274729013443, "learning_rate": 9.378093758526316e-05, "loss": 0.0405, "step": 32390 }, { "epoch": 0.6474561368450502, "grad_norm": 0.17704905569553375, "learning_rate": 9.37759452081941e-05, "loss": 0.0476, "step": 32400 }, { "epoch": 0.6476559689860517, "grad_norm": 0.15816693007946014, "learning_rate": 9.37709509610971e-05, "loss": 0.0337, "step": 32410 }, { "epoch": 0.6478558011270533, "grad_norm": 0.1471642702817917, "learning_rate": 9.376595484418548e-05, "loss": 0.036, "step": 32420 }, { "epoch": 0.6480556332680548, "grad_norm": 0.1077786311507225, "learning_rate": 9.376095685767267e-05, "loss": 0.035, "step": 32430 }, { "epoch": 0.6482554654090564, "grad_norm": 0.15503224730491638, "learning_rate": 9.375595700177217e-05, "loss": 0.0443, "step": 32440 }, { "epoch": 0.6484552975500579, "grad_norm": 0.14276400208473206, "learning_rate": 9.375095527669758e-05, "loss": 0.0356, "step": 32450 }, { "epoch": 0.6486551296910595, "grad_norm": 0.1749453842639923, "learning_rate": 9.374595168266253e-05, "loss": 0.0509, "step": 32460 }, { "epoch": 0.648854961832061, "grad_norm": 0.1382509022951126, "learning_rate": 9.37409462198808e-05, "loss": 0.0387, "step": 32470 }, { "epoch": 0.6490547939730626, "grad_norm": 0.1463388353586197, "learning_rate": 9.37359388885662e-05, "loss": 0.0467, "step": 32480 }, { "epoch": 0.6492546261140641, "grad_norm": 0.12948200106620789, "learning_rate": 9.373092968893265e-05, "loss": 0.0432, "step": 32490 }, { "epoch": 0.6494544582550658, "grad_norm": 0.12028687447309494, "learning_rate": 9.372591862119411e-05, "loss": 0.0336, "step": 32500 }, { "epoch": 0.6496542903960673, "grad_norm": 0.14614878594875336, "learning_rate": 9.372090568556465e-05, "loss": 0.0432, "step": 32510 }, { "epoch": 0.6498541225370689, "grad_norm": 0.17441321909427643, "learning_rate": 9.371589088225843e-05, "loss": 0.048, "step": 32520 }, { "epoch": 0.6500539546780704, "grad_norm": 0.13996772468090057, "learning_rate": 9.371087421148964e-05, "loss": 0.0421, "step": 32530 }, { "epoch": 0.650253786819072, "grad_norm": 0.15388429164886475, "learning_rate": 9.370585567347261e-05, "loss": 0.0327, "step": 32540 }, { "epoch": 0.6504536189600736, "grad_norm": 0.1492241770029068, "learning_rate": 9.370083526842173e-05, "loss": 0.0347, "step": 32550 }, { "epoch": 0.6506534511010751, "grad_norm": 0.11678627133369446, "learning_rate": 9.369581299655143e-05, "loss": 0.0414, "step": 32560 }, { "epoch": 0.6508532832420767, "grad_norm": 0.1658342480659485, "learning_rate": 9.369078885807627e-05, "loss": 0.0498, "step": 32570 }, { "epoch": 0.6510531153830782, "grad_norm": 0.14558041095733643, "learning_rate": 9.36857628532109e-05, "loss": 0.0336, "step": 32580 }, { "epoch": 0.6512529475240798, "grad_norm": 0.11047597229480743, "learning_rate": 9.368073498216997e-05, "loss": 0.0383, "step": 32590 }, { "epoch": 0.6514527796650813, "grad_norm": 0.16374674439430237, "learning_rate": 9.36757052451683e-05, "loss": 0.0419, "step": 32600 }, { "epoch": 0.6516526118060829, "grad_norm": 0.12098772823810577, "learning_rate": 9.367067364242072e-05, "loss": 0.0419, "step": 32610 }, { "epoch": 0.6518524439470844, "grad_norm": 0.13666732609272003, "learning_rate": 9.366564017414219e-05, "loss": 0.0411, "step": 32620 }, { "epoch": 0.652052276088086, "grad_norm": 0.13219045102596283, "learning_rate": 9.366060484054773e-05, "loss": 0.0411, "step": 32630 }, { "epoch": 0.6522521082290875, "grad_norm": 0.1539595127105713, "learning_rate": 9.365556764185245e-05, "loss": 0.0363, "step": 32640 }, { "epoch": 0.6524519403700891, "grad_norm": 0.12382572889328003, "learning_rate": 9.36505285782715e-05, "loss": 0.0376, "step": 32650 }, { "epoch": 0.6526517725110906, "grad_norm": 0.12671196460723877, "learning_rate": 9.364548765002017e-05, "loss": 0.049, "step": 32660 }, { "epoch": 0.6528516046520922, "grad_norm": 0.14248797297477722, "learning_rate": 9.364044485731378e-05, "loss": 0.0463, "step": 32670 }, { "epoch": 0.6530514367930939, "grad_norm": 0.11943532526493073, "learning_rate": 9.363540020036777e-05, "loss": 0.0316, "step": 32680 }, { "epoch": 0.6532512689340954, "grad_norm": 0.12670494616031647, "learning_rate": 9.363035367939761e-05, "loss": 0.0288, "step": 32690 }, { "epoch": 0.653451101075097, "grad_norm": 0.10597623884677887, "learning_rate": 9.36253052946189e-05, "loss": 0.0407, "step": 32700 }, { "epoch": 0.6536509332160985, "grad_norm": 0.11907882988452911, "learning_rate": 9.362025504624729e-05, "loss": 0.0422, "step": 32710 }, { "epoch": 0.6538507653571001, "grad_norm": 0.20205166935920715, "learning_rate": 9.361520293449851e-05, "loss": 0.0438, "step": 32720 }, { "epoch": 0.6540505974981016, "grad_norm": 0.1615743786096573, "learning_rate": 9.361014895958839e-05, "loss": 0.0396, "step": 32730 }, { "epoch": 0.6542504296391032, "grad_norm": 0.1312638819217682, "learning_rate": 9.360509312173283e-05, "loss": 0.0409, "step": 32740 }, { "epoch": 0.6544502617801047, "grad_norm": 0.12253518402576447, "learning_rate": 9.360003542114777e-05, "loss": 0.0427, "step": 32750 }, { "epoch": 0.6546500939211063, "grad_norm": 0.19975140690803528, "learning_rate": 9.359497585804931e-05, "loss": 0.0435, "step": 32760 }, { "epoch": 0.6548499260621078, "grad_norm": 0.15403762459754944, "learning_rate": 9.358991443265357e-05, "loss": 0.0426, "step": 32770 }, { "epoch": 0.6550497582031094, "grad_norm": 0.12897419929504395, "learning_rate": 9.358485114517676e-05, "loss": 0.0372, "step": 32780 }, { "epoch": 0.6552495903441109, "grad_norm": 0.14020346105098724, "learning_rate": 9.357978599583518e-05, "loss": 0.0444, "step": 32790 }, { "epoch": 0.6554494224851125, "grad_norm": 0.13809987902641296, "learning_rate": 9.35747189848452e-05, "loss": 0.0372, "step": 32800 }, { "epoch": 0.6556492546261141, "grad_norm": 0.17884893715381622, "learning_rate": 9.356965011242327e-05, "loss": 0.0431, "step": 32810 }, { "epoch": 0.6558490867671156, "grad_norm": 0.1449095755815506, "learning_rate": 9.356457937878593e-05, "loss": 0.0426, "step": 32820 }, { "epoch": 0.6560489189081172, "grad_norm": 0.13320055603981018, "learning_rate": 9.355950678414979e-05, "loss": 0.0316, "step": 32830 }, { "epoch": 0.6562487510491187, "grad_norm": 0.12856262922286987, "learning_rate": 9.355443232873154e-05, "loss": 0.038, "step": 32840 }, { "epoch": 0.6564485831901203, "grad_norm": 0.16133694350719452, "learning_rate": 9.354935601274795e-05, "loss": 0.0356, "step": 32850 }, { "epoch": 0.6566484153311218, "grad_norm": 0.13960064947605133, "learning_rate": 9.354427783641589e-05, "loss": 0.0324, "step": 32860 }, { "epoch": 0.6568482474721234, "grad_norm": 0.13204407691955566, "learning_rate": 9.353919779995224e-05, "loss": 0.0393, "step": 32870 }, { "epoch": 0.657048079613125, "grad_norm": 0.15496699512004852, "learning_rate": 9.353411590357407e-05, "loss": 0.0386, "step": 32880 }, { "epoch": 0.6572479117541266, "grad_norm": 0.1272636353969574, "learning_rate": 9.352903214749844e-05, "loss": 0.0433, "step": 32890 }, { "epoch": 0.6574477438951281, "grad_norm": 0.10447163879871368, "learning_rate": 9.352394653194253e-05, "loss": 0.0278, "step": 32900 }, { "epoch": 0.6576475760361297, "grad_norm": 0.17764581739902496, "learning_rate": 9.351885905712357e-05, "loss": 0.0363, "step": 32910 }, { "epoch": 0.6578474081771312, "grad_norm": 0.17200036346912384, "learning_rate": 9.35137697232589e-05, "loss": 0.0447, "step": 32920 }, { "epoch": 0.6580472403181328, "grad_norm": 0.16317196190357208, "learning_rate": 9.350867853056591e-05, "loss": 0.0348, "step": 32930 }, { "epoch": 0.6582470724591343, "grad_norm": 0.16994625329971313, "learning_rate": 9.350358547926212e-05, "loss": 0.0444, "step": 32940 }, { "epoch": 0.6584469046001359, "grad_norm": 0.1643165647983551, "learning_rate": 9.349849056956508e-05, "loss": 0.0423, "step": 32950 }, { "epoch": 0.6586467367411375, "grad_norm": 0.13572652637958527, "learning_rate": 9.349339380169242e-05, "loss": 0.0392, "step": 32960 }, { "epoch": 0.658846568882139, "grad_norm": 0.14745725691318512, "learning_rate": 9.348829517586187e-05, "loss": 0.0462, "step": 32970 }, { "epoch": 0.6590464010231406, "grad_norm": 0.12358028441667557, "learning_rate": 9.348319469229126e-05, "loss": 0.0424, "step": 32980 }, { "epoch": 0.6592462331641421, "grad_norm": 0.14507867395877838, "learning_rate": 9.347809235119844e-05, "loss": 0.0433, "step": 32990 }, { "epoch": 0.6594460653051437, "grad_norm": 0.14245425164699554, "learning_rate": 9.34729881528014e-05, "loss": 0.0375, "step": 33000 }, { "epoch": 0.6596458974461452, "grad_norm": 0.1286691129207611, "learning_rate": 9.346788209731815e-05, "loss": 0.0293, "step": 33010 }, { "epoch": 0.6598457295871468, "grad_norm": 0.16478486359119415, "learning_rate": 9.346277418496685e-05, "loss": 0.0433, "step": 33020 }, { "epoch": 0.6600455617281483, "grad_norm": 0.13134843111038208, "learning_rate": 9.345766441596566e-05, "loss": 0.0376, "step": 33030 }, { "epoch": 0.6602453938691499, "grad_norm": 0.13091135025024414, "learning_rate": 9.345255279053288e-05, "loss": 0.0408, "step": 33040 }, { "epoch": 0.6604452260101514, "grad_norm": 0.14722439646720886, "learning_rate": 9.344743930888689e-05, "loss": 0.0442, "step": 33050 }, { "epoch": 0.660645058151153, "grad_norm": 0.1961469054222107, "learning_rate": 9.34423239712461e-05, "loss": 0.0311, "step": 33060 }, { "epoch": 0.6608448902921545, "grad_norm": 0.14003150165081024, "learning_rate": 9.343720677782904e-05, "loss": 0.0406, "step": 33070 }, { "epoch": 0.6610447224331562, "grad_norm": 0.16267427802085876, "learning_rate": 9.34320877288543e-05, "loss": 0.0378, "step": 33080 }, { "epoch": 0.6612445545741578, "grad_norm": 0.10454929620027542, "learning_rate": 9.342696682454055e-05, "loss": 0.0486, "step": 33090 }, { "epoch": 0.6614443867151593, "grad_norm": 0.17050690948963165, "learning_rate": 9.342184406510657e-05, "loss": 0.0384, "step": 33100 }, { "epoch": 0.6616442188561609, "grad_norm": 0.1452067494392395, "learning_rate": 9.341671945077118e-05, "loss": 0.0382, "step": 33110 }, { "epoch": 0.6618440509971624, "grad_norm": 0.12773533165454865, "learning_rate": 9.341159298175328e-05, "loss": 0.0348, "step": 33120 }, { "epoch": 0.662043883138164, "grad_norm": 0.1087999939918518, "learning_rate": 9.340646465827188e-05, "loss": 0.0402, "step": 33130 }, { "epoch": 0.6622437152791655, "grad_norm": 0.18621514737606049, "learning_rate": 9.340133448054605e-05, "loss": 0.0406, "step": 33140 }, { "epoch": 0.6624435474201671, "grad_norm": 0.13520805537700653, "learning_rate": 9.339620244879495e-05, "loss": 0.0467, "step": 33150 }, { "epoch": 0.6626433795611686, "grad_norm": 0.13390734791755676, "learning_rate": 9.339106856323781e-05, "loss": 0.0462, "step": 33160 }, { "epoch": 0.6628432117021702, "grad_norm": 0.12314392626285553, "learning_rate": 9.338593282409392e-05, "loss": 0.0434, "step": 33170 }, { "epoch": 0.6630430438431717, "grad_norm": 0.12879902124404907, "learning_rate": 9.338079523158267e-05, "loss": 0.0322, "step": 33180 }, { "epoch": 0.6632428759841733, "grad_norm": 0.1651608794927597, "learning_rate": 9.337565578592355e-05, "loss": 0.044, "step": 33190 }, { "epoch": 0.6634427081251748, "grad_norm": 0.1898788958787918, "learning_rate": 9.337051448733611e-05, "loss": 0.0352, "step": 33200 }, { "epoch": 0.6636425402661764, "grad_norm": 0.15186849236488342, "learning_rate": 9.336537133603996e-05, "loss": 0.0415, "step": 33210 }, { "epoch": 0.6638423724071779, "grad_norm": 0.13188323378562927, "learning_rate": 9.33602263322548e-05, "loss": 0.0376, "step": 33220 }, { "epoch": 0.6640422045481795, "grad_norm": 0.12513059377670288, "learning_rate": 9.335507947620044e-05, "loss": 0.0363, "step": 33230 }, { "epoch": 0.6642420366891811, "grad_norm": 0.1577702760696411, "learning_rate": 9.334993076809673e-05, "loss": 0.0458, "step": 33240 }, { "epoch": 0.6644418688301826, "grad_norm": 0.15513086318969727, "learning_rate": 9.334478020816361e-05, "loss": 0.0418, "step": 33250 }, { "epoch": 0.6646417009711842, "grad_norm": 0.14115969836711884, "learning_rate": 9.333962779662111e-05, "loss": 0.0347, "step": 33260 }, { "epoch": 0.6648415331121857, "grad_norm": 0.11206420511007309, "learning_rate": 9.333447353368932e-05, "loss": 0.0375, "step": 33270 }, { "epoch": 0.6650413652531874, "grad_norm": 0.11306747794151306, "learning_rate": 9.332931741958843e-05, "loss": 0.036, "step": 33280 }, { "epoch": 0.6652411973941889, "grad_norm": 0.12665912508964539, "learning_rate": 9.332415945453871e-05, "loss": 0.0384, "step": 33290 }, { "epoch": 0.6654410295351905, "grad_norm": 0.13547196984291077, "learning_rate": 9.331899963876047e-05, "loss": 0.0455, "step": 33300 }, { "epoch": 0.665640861676192, "grad_norm": 0.10675729811191559, "learning_rate": 9.331383797247417e-05, "loss": 0.045, "step": 33310 }, { "epoch": 0.6658406938171936, "grad_norm": 0.1408648043870926, "learning_rate": 9.330867445590026e-05, "loss": 0.0312, "step": 33320 }, { "epoch": 0.6660405259581951, "grad_norm": 0.13093498349189758, "learning_rate": 9.330350908925934e-05, "loss": 0.0398, "step": 33330 }, { "epoch": 0.6662403580991967, "grad_norm": 0.14969877898693085, "learning_rate": 9.329834187277206e-05, "loss": 0.0401, "step": 33340 }, { "epoch": 0.6664401902401982, "grad_norm": 0.1355314403772354, "learning_rate": 9.329317280665916e-05, "loss": 0.04, "step": 33350 }, { "epoch": 0.6666400223811998, "grad_norm": 0.14663413166999817, "learning_rate": 9.328800189114146e-05, "loss": 0.0345, "step": 33360 }, { "epoch": 0.6668398545222014, "grad_norm": 0.10437164455652237, "learning_rate": 9.328282912643985e-05, "loss": 0.0336, "step": 33370 }, { "epoch": 0.6670396866632029, "grad_norm": 0.15774469077587128, "learning_rate": 9.327765451277527e-05, "loss": 0.0445, "step": 33380 }, { "epoch": 0.6672395188042045, "grad_norm": 0.1534651219844818, "learning_rate": 9.32724780503688e-05, "loss": 0.0404, "step": 33390 }, { "epoch": 0.667439350945206, "grad_norm": 0.17046315968036652, "learning_rate": 9.326729973944157e-05, "loss": 0.0488, "step": 33400 }, { "epoch": 0.6676391830862076, "grad_norm": 0.17131510376930237, "learning_rate": 9.326211958021475e-05, "loss": 0.0376, "step": 33410 }, { "epoch": 0.6678390152272091, "grad_norm": 0.20749825239181519, "learning_rate": 9.32569375729097e-05, "loss": 0.0328, "step": 33420 }, { "epoch": 0.6680388473682107, "grad_norm": 0.13588832318782806, "learning_rate": 9.325175371774771e-05, "loss": 0.0322, "step": 33430 }, { "epoch": 0.6682386795092122, "grad_norm": 0.15911568701267242, "learning_rate": 9.324656801495027e-05, "loss": 0.0484, "step": 33440 }, { "epoch": 0.6684385116502138, "grad_norm": 0.1279090940952301, "learning_rate": 9.32413804647389e-05, "loss": 0.0384, "step": 33450 }, { "epoch": 0.6686383437912153, "grad_norm": 0.12234438210725784, "learning_rate": 9.323619106733517e-05, "loss": 0.0302, "step": 33460 }, { "epoch": 0.668838175932217, "grad_norm": 0.14177894592285156, "learning_rate": 9.323099982296079e-05, "loss": 0.0382, "step": 33470 }, { "epoch": 0.6690380080732184, "grad_norm": 0.11784283071756363, "learning_rate": 9.322580673183754e-05, "loss": 0.0321, "step": 33480 }, { "epoch": 0.6692378402142201, "grad_norm": 0.1395537555217743, "learning_rate": 9.322061179418722e-05, "loss": 0.0357, "step": 33490 }, { "epoch": 0.6694376723552217, "grad_norm": 0.14206120371818542, "learning_rate": 9.321541501023175e-05, "loss": 0.0351, "step": 33500 }, { "epoch": 0.6696375044962232, "grad_norm": 0.11916078627109528, "learning_rate": 9.321021638019317e-05, "loss": 0.0409, "step": 33510 }, { "epoch": 0.6698373366372248, "grad_norm": 0.14171773195266724, "learning_rate": 9.320501590429349e-05, "loss": 0.042, "step": 33520 }, { "epoch": 0.6700371687782263, "grad_norm": 0.1427420824766159, "learning_rate": 9.319981358275492e-05, "loss": 0.0506, "step": 33530 }, { "epoch": 0.6702370009192279, "grad_norm": 0.11429426074028015, "learning_rate": 9.319460941579967e-05, "loss": 0.0329, "step": 33540 }, { "epoch": 0.6704368330602294, "grad_norm": 0.13344980776309967, "learning_rate": 9.318940340365005e-05, "loss": 0.035, "step": 33550 }, { "epoch": 0.670636665201231, "grad_norm": 0.15199516713619232, "learning_rate": 9.318419554652848e-05, "loss": 0.0415, "step": 33560 }, { "epoch": 0.6708364973422325, "grad_norm": 0.1250184029340744, "learning_rate": 9.317898584465738e-05, "loss": 0.0432, "step": 33570 }, { "epoch": 0.6710363294832341, "grad_norm": 0.14875061810016632, "learning_rate": 9.317377429825935e-05, "loss": 0.0381, "step": 33580 }, { "epoch": 0.6712361616242356, "grad_norm": 0.14613504707813263, "learning_rate": 9.316856090755699e-05, "loss": 0.0407, "step": 33590 }, { "epoch": 0.6714359937652372, "grad_norm": 0.12515988945960999, "learning_rate": 9.3163345672773e-05, "loss": 0.034, "step": 33600 }, { "epoch": 0.6716358259062387, "grad_norm": 0.12581737339496613, "learning_rate": 9.31581285941302e-05, "loss": 0.0414, "step": 33610 }, { "epoch": 0.6718356580472403, "grad_norm": 0.148957759141922, "learning_rate": 9.31529096718514e-05, "loss": 0.0415, "step": 33620 }, { "epoch": 0.6720354901882418, "grad_norm": 0.12694716453552246, "learning_rate": 9.31476889061596e-05, "loss": 0.0446, "step": 33630 }, { "epoch": 0.6722353223292434, "grad_norm": 0.12573985755443573, "learning_rate": 9.314246629727779e-05, "loss": 0.0387, "step": 33640 }, { "epoch": 0.672435154470245, "grad_norm": 0.1393265724182129, "learning_rate": 9.313724184542906e-05, "loss": 0.0369, "step": 33650 }, { "epoch": 0.6726349866112465, "grad_norm": 0.1342366337776184, "learning_rate": 9.313201555083662e-05, "loss": 0.0399, "step": 33660 }, { "epoch": 0.6728348187522482, "grad_norm": 0.14083003997802734, "learning_rate": 9.312678741372369e-05, "loss": 0.0414, "step": 33670 }, { "epoch": 0.6730346508932497, "grad_norm": 0.14947734773159027, "learning_rate": 9.312155743431364e-05, "loss": 0.0388, "step": 33680 }, { "epoch": 0.6732344830342513, "grad_norm": 0.10162945091724396, "learning_rate": 9.311632561282988e-05, "loss": 0.037, "step": 33690 }, { "epoch": 0.6734343151752528, "grad_norm": 0.24474261701107025, "learning_rate": 9.311109194949589e-05, "loss": 0.0367, "step": 33700 }, { "epoch": 0.6736341473162544, "grad_norm": 0.14719536900520325, "learning_rate": 9.310585644453524e-05, "loss": 0.0391, "step": 33710 }, { "epoch": 0.6738339794572559, "grad_norm": 0.21162423491477966, "learning_rate": 9.310061909817161e-05, "loss": 0.0331, "step": 33720 }, { "epoch": 0.6740338115982575, "grad_norm": 0.1316078156232834, "learning_rate": 9.309537991062872e-05, "loss": 0.0349, "step": 33730 }, { "epoch": 0.674233643739259, "grad_norm": 0.1056961864233017, "learning_rate": 9.309013888213034e-05, "loss": 0.0386, "step": 33740 }, { "epoch": 0.6744334758802606, "grad_norm": 0.1474645435810089, "learning_rate": 9.30848960129004e-05, "loss": 0.0441, "step": 33750 }, { "epoch": 0.6746333080212621, "grad_norm": 0.1415363997220993, "learning_rate": 9.307965130316284e-05, "loss": 0.0419, "step": 33760 }, { "epoch": 0.6748331401622637, "grad_norm": 0.1347956359386444, "learning_rate": 9.307440475314173e-05, "loss": 0.0438, "step": 33770 }, { "epoch": 0.6750329723032653, "grad_norm": 0.14662884175777435, "learning_rate": 9.306915636306117e-05, "loss": 0.0421, "step": 33780 }, { "epoch": 0.6752328044442668, "grad_norm": 0.1497027426958084, "learning_rate": 9.306390613314538e-05, "loss": 0.055, "step": 33790 }, { "epoch": 0.6754326365852684, "grad_norm": 0.1519414782524109, "learning_rate": 9.305865406361862e-05, "loss": 0.0465, "step": 33800 }, { "epoch": 0.6756324687262699, "grad_norm": 0.16993486881256104, "learning_rate": 9.305340015470527e-05, "loss": 0.0438, "step": 33810 }, { "epoch": 0.6758323008672715, "grad_norm": 0.15547555685043335, "learning_rate": 9.304814440662977e-05, "loss": 0.0381, "step": 33820 }, { "epoch": 0.676032133008273, "grad_norm": 0.14265039563179016, "learning_rate": 9.30428868196166e-05, "loss": 0.0347, "step": 33830 }, { "epoch": 0.6762319651492746, "grad_norm": 0.13657891750335693, "learning_rate": 9.303762739389036e-05, "loss": 0.0371, "step": 33840 }, { "epoch": 0.6764317972902761, "grad_norm": 0.14299222826957703, "learning_rate": 9.303236612967578e-05, "loss": 0.0307, "step": 33850 }, { "epoch": 0.6766316294312777, "grad_norm": 0.11272513121366501, "learning_rate": 9.302710302719756e-05, "loss": 0.0447, "step": 33860 }, { "epoch": 0.6768314615722792, "grad_norm": 0.13527649641036987, "learning_rate": 9.302183808668053e-05, "loss": 0.036, "step": 33870 }, { "epoch": 0.6770312937132809, "grad_norm": 0.12751202285289764, "learning_rate": 9.301657130834962e-05, "loss": 0.0465, "step": 33880 }, { "epoch": 0.6772311258542824, "grad_norm": 0.16368843615055084, "learning_rate": 9.301130269242982e-05, "loss": 0.0298, "step": 33890 }, { "epoch": 0.677430957995284, "grad_norm": 0.16197089850902557, "learning_rate": 9.300603223914615e-05, "loss": 0.039, "step": 33900 }, { "epoch": 0.6776307901362855, "grad_norm": 0.18883788585662842, "learning_rate": 9.300075994872382e-05, "loss": 0.0485, "step": 33910 }, { "epoch": 0.6778306222772871, "grad_norm": 0.16491127014160156, "learning_rate": 9.2995485821388e-05, "loss": 0.0353, "step": 33920 }, { "epoch": 0.6780304544182887, "grad_norm": 0.13090680539608002, "learning_rate": 9.299020985736402e-05, "loss": 0.0507, "step": 33930 }, { "epoch": 0.6782302865592902, "grad_norm": 0.12110438942909241, "learning_rate": 9.298493205687725e-05, "loss": 0.0371, "step": 33940 }, { "epoch": 0.6784301187002918, "grad_norm": 0.16219203174114227, "learning_rate": 9.297965242015315e-05, "loss": 0.0322, "step": 33950 }, { "epoch": 0.6786299508412933, "grad_norm": 0.1483324021100998, "learning_rate": 9.297437094741724e-05, "loss": 0.0398, "step": 33960 }, { "epoch": 0.6788297829822949, "grad_norm": 0.10871393978595734, "learning_rate": 9.296908763889516e-05, "loss": 0.041, "step": 33970 }, { "epoch": 0.6790296151232964, "grad_norm": 0.12060420215129852, "learning_rate": 9.29638024948126e-05, "loss": 0.0437, "step": 33980 }, { "epoch": 0.679229447264298, "grad_norm": 0.14820222556591034, "learning_rate": 9.295851551539532e-05, "loss": 0.0417, "step": 33990 }, { "epoch": 0.6794292794052995, "grad_norm": 0.15012003481388092, "learning_rate": 9.295322670086917e-05, "loss": 0.0538, "step": 34000 }, { "epoch": 0.6796291115463011, "grad_norm": 0.17732909321784973, "learning_rate": 9.294793605146009e-05, "loss": 0.0401, "step": 34010 }, { "epoch": 0.6798289436873026, "grad_norm": 0.17593763768672943, "learning_rate": 9.294264356739407e-05, "loss": 0.0374, "step": 34020 }, { "epoch": 0.6800287758283042, "grad_norm": 0.1577867567539215, "learning_rate": 9.293734924889721e-05, "loss": 0.0404, "step": 34030 }, { "epoch": 0.6802286079693057, "grad_norm": 0.11794978380203247, "learning_rate": 9.293205309619566e-05, "loss": 0.0398, "step": 34040 }, { "epoch": 0.6804284401103073, "grad_norm": 0.10020491480827332, "learning_rate": 9.292675510951568e-05, "loss": 0.0463, "step": 34050 }, { "epoch": 0.680628272251309, "grad_norm": 0.18156783282756805, "learning_rate": 9.292145528908357e-05, "loss": 0.0354, "step": 34060 }, { "epoch": 0.6808281043923105, "grad_norm": 0.14168371260166168, "learning_rate": 9.291615363512576e-05, "loss": 0.0353, "step": 34070 }, { "epoch": 0.6810279365333121, "grad_norm": 0.14851246774196625, "learning_rate": 9.29108501478687e-05, "loss": 0.0378, "step": 34080 }, { "epoch": 0.6812277686743136, "grad_norm": 0.1309136599302292, "learning_rate": 9.290554482753895e-05, "loss": 0.0348, "step": 34090 }, { "epoch": 0.6814276008153152, "grad_norm": 0.10967034101486206, "learning_rate": 9.290023767436314e-05, "loss": 0.0417, "step": 34100 }, { "epoch": 0.6816274329563167, "grad_norm": 0.17796102166175842, "learning_rate": 9.289492868856799e-05, "loss": 0.0393, "step": 34110 }, { "epoch": 0.6818272650973183, "grad_norm": 0.18336069583892822, "learning_rate": 9.288961787038029e-05, "loss": 0.0407, "step": 34120 }, { "epoch": 0.6820270972383198, "grad_norm": 0.12247339636087418, "learning_rate": 9.288430522002689e-05, "loss": 0.0326, "step": 34130 }, { "epoch": 0.6822269293793214, "grad_norm": 0.18286441266536713, "learning_rate": 9.287899073773476e-05, "loss": 0.0485, "step": 34140 }, { "epoch": 0.6824267615203229, "grad_norm": 0.17908120155334473, "learning_rate": 9.287367442373092e-05, "loss": 0.0485, "step": 34150 }, { "epoch": 0.6826265936613245, "grad_norm": 0.14464512467384338, "learning_rate": 9.286835627824247e-05, "loss": 0.0412, "step": 34160 }, { "epoch": 0.682826425802326, "grad_norm": 0.1574501097202301, "learning_rate": 9.28630363014966e-05, "loss": 0.0346, "step": 34170 }, { "epoch": 0.6830262579433276, "grad_norm": 0.11749158799648285, "learning_rate": 9.285771449372054e-05, "loss": 0.0323, "step": 34180 }, { "epoch": 0.6832260900843292, "grad_norm": 0.15074801445007324, "learning_rate": 9.285239085514167e-05, "loss": 0.0441, "step": 34190 }, { "epoch": 0.6834259222253307, "grad_norm": 0.15463972091674805, "learning_rate": 9.284706538598737e-05, "loss": 0.031, "step": 34200 }, { "epoch": 0.6836257543663323, "grad_norm": 0.1256045550107956, "learning_rate": 9.284173808648515e-05, "loss": 0.0411, "step": 34210 }, { "epoch": 0.6838255865073338, "grad_norm": 0.13506120443344116, "learning_rate": 9.28364089568626e-05, "loss": 0.0429, "step": 34220 }, { "epoch": 0.6840254186483354, "grad_norm": 0.16377919912338257, "learning_rate": 9.283107799734734e-05, "loss": 0.0427, "step": 34230 }, { "epoch": 0.6842252507893369, "grad_norm": 0.12193906307220459, "learning_rate": 9.282574520816711e-05, "loss": 0.0423, "step": 34240 }, { "epoch": 0.6844250829303385, "grad_norm": 0.12329705059528351, "learning_rate": 9.282041058954971e-05, "loss": 0.0397, "step": 34250 }, { "epoch": 0.68462491507134, "grad_norm": 0.1538211554288864, "learning_rate": 9.281507414172306e-05, "loss": 0.0537, "step": 34260 }, { "epoch": 0.6848247472123417, "grad_norm": 0.15524010360240936, "learning_rate": 9.280973586491508e-05, "loss": 0.0501, "step": 34270 }, { "epoch": 0.6850245793533432, "grad_norm": 0.10930430144071579, "learning_rate": 9.280439575935381e-05, "loss": 0.0325, "step": 34280 }, { "epoch": 0.6852244114943448, "grad_norm": 0.13419939577579498, "learning_rate": 9.279905382526741e-05, "loss": 0.0364, "step": 34290 }, { "epoch": 0.6854242436353463, "grad_norm": 0.1303047239780426, "learning_rate": 9.279371006288404e-05, "loss": 0.0429, "step": 34300 }, { "epoch": 0.6856240757763479, "grad_norm": 0.12079041451215744, "learning_rate": 9.2788364472432e-05, "loss": 0.0406, "step": 34310 }, { "epoch": 0.6858239079173494, "grad_norm": 0.17294226586818695, "learning_rate": 9.278301705413964e-05, "loss": 0.0447, "step": 34320 }, { "epoch": 0.686023740058351, "grad_norm": 0.1440272480249405, "learning_rate": 9.277766780823538e-05, "loss": 0.0359, "step": 34330 }, { "epoch": 0.6862235721993526, "grad_norm": 0.1465797871351242, "learning_rate": 9.277231673494773e-05, "loss": 0.0378, "step": 34340 }, { "epoch": 0.6864234043403541, "grad_norm": 0.13895289599895477, "learning_rate": 9.276696383450529e-05, "loss": 0.0374, "step": 34350 }, { "epoch": 0.6866232364813557, "grad_norm": 0.16098320484161377, "learning_rate": 9.276160910713673e-05, "loss": 0.0476, "step": 34360 }, { "epoch": 0.6868230686223572, "grad_norm": 0.16807588934898376, "learning_rate": 9.275625255307077e-05, "loss": 0.0391, "step": 34370 }, { "epoch": 0.6870229007633588, "grad_norm": 0.2065885066986084, "learning_rate": 9.275089417253626e-05, "loss": 0.0381, "step": 34380 }, { "epoch": 0.6872227329043603, "grad_norm": 0.16707219183444977, "learning_rate": 9.274553396576208e-05, "loss": 0.0339, "step": 34390 }, { "epoch": 0.6874225650453619, "grad_norm": 0.12502604722976685, "learning_rate": 9.274017193297721e-05, "loss": 0.0454, "step": 34400 }, { "epoch": 0.6876223971863634, "grad_norm": 0.15345922112464905, "learning_rate": 9.273480807441071e-05, "loss": 0.0409, "step": 34410 }, { "epoch": 0.687822229327365, "grad_norm": 0.14060436189174652, "learning_rate": 9.272944239029173e-05, "loss": 0.0401, "step": 34420 }, { "epoch": 0.6880220614683665, "grad_norm": 0.14666511118412018, "learning_rate": 9.272407488084945e-05, "loss": 0.0407, "step": 34430 }, { "epoch": 0.6882218936093681, "grad_norm": 0.1299176961183548, "learning_rate": 9.271870554631319e-05, "loss": 0.0358, "step": 34440 }, { "epoch": 0.6884217257503696, "grad_norm": 0.1703968048095703, "learning_rate": 9.27133343869123e-05, "loss": 0.0421, "step": 34450 }, { "epoch": 0.6886215578913712, "grad_norm": 0.13406436145305634, "learning_rate": 9.270796140287625e-05, "loss": 0.0359, "step": 34460 }, { "epoch": 0.6888213900323729, "grad_norm": 0.166203111410141, "learning_rate": 9.270258659443452e-05, "loss": 0.0354, "step": 34470 }, { "epoch": 0.6890212221733744, "grad_norm": 0.15057355165481567, "learning_rate": 9.269720996181674e-05, "loss": 0.0434, "step": 34480 }, { "epoch": 0.689221054314376, "grad_norm": 0.13151709735393524, "learning_rate": 9.26918315052526e-05, "loss": 0.0492, "step": 34490 }, { "epoch": 0.6894208864553775, "grad_norm": 0.16253253817558289, "learning_rate": 9.268645122497184e-05, "loss": 0.0349, "step": 34500 }, { "epoch": 0.6896207185963791, "grad_norm": 0.15234078466892242, "learning_rate": 9.268106912120431e-05, "loss": 0.0416, "step": 34510 }, { "epoch": 0.6898205507373806, "grad_norm": 0.12226312607526779, "learning_rate": 9.267568519417991e-05, "loss": 0.0398, "step": 34520 }, { "epoch": 0.6900203828783822, "grad_norm": 0.15237659215927124, "learning_rate": 9.267029944412862e-05, "loss": 0.0396, "step": 34530 }, { "epoch": 0.6902202150193837, "grad_norm": 0.17251349985599518, "learning_rate": 9.266491187128054e-05, "loss": 0.0439, "step": 34540 }, { "epoch": 0.6904200471603853, "grad_norm": 0.13516266644001007, "learning_rate": 9.26595224758658e-05, "loss": 0.0336, "step": 34550 }, { "epoch": 0.6906198793013868, "grad_norm": 0.13968461751937866, "learning_rate": 9.265413125811465e-05, "loss": 0.0481, "step": 34560 }, { "epoch": 0.6908197114423884, "grad_norm": 0.13648928701877594, "learning_rate": 9.264873821825734e-05, "loss": 0.0409, "step": 34570 }, { "epoch": 0.6910195435833899, "grad_norm": 0.1505601853132248, "learning_rate": 9.26433433565243e-05, "loss": 0.0377, "step": 34580 }, { "epoch": 0.6912193757243915, "grad_norm": 0.1336437612771988, "learning_rate": 9.263794667314596e-05, "loss": 0.0448, "step": 34590 }, { "epoch": 0.691419207865393, "grad_norm": 0.14432162046432495, "learning_rate": 9.263254816835285e-05, "loss": 0.0436, "step": 34600 }, { "epoch": 0.6916190400063946, "grad_norm": 0.13817928731441498, "learning_rate": 9.262714784237562e-05, "loss": 0.0384, "step": 34610 }, { "epoch": 0.6918188721473962, "grad_norm": 0.16892851889133453, "learning_rate": 9.262174569544495e-05, "loss": 0.0401, "step": 34620 }, { "epoch": 0.6920187042883977, "grad_norm": 0.1438208818435669, "learning_rate": 9.26163417277916e-05, "loss": 0.0316, "step": 34630 }, { "epoch": 0.6922185364293993, "grad_norm": 0.14187942445278168, "learning_rate": 9.26109359396464e-05, "loss": 0.0379, "step": 34640 }, { "epoch": 0.6924183685704008, "grad_norm": 0.14031270146369934, "learning_rate": 9.260552833124031e-05, "loss": 0.0344, "step": 34650 }, { "epoch": 0.6926182007114025, "grad_norm": 0.13967685401439667, "learning_rate": 9.260011890280434e-05, "loss": 0.0514, "step": 34660 }, { "epoch": 0.692818032852404, "grad_norm": 0.1286783516407013, "learning_rate": 9.259470765456952e-05, "loss": 0.046, "step": 34670 }, { "epoch": 0.6930178649934056, "grad_norm": 0.14015083014965057, "learning_rate": 9.258929458676704e-05, "loss": 0.0394, "step": 34680 }, { "epoch": 0.6932176971344071, "grad_norm": 0.12687860429286957, "learning_rate": 9.258387969962814e-05, "loss": 0.0411, "step": 34690 }, { "epoch": 0.6934175292754087, "grad_norm": 0.14383485913276672, "learning_rate": 9.257846299338413e-05, "loss": 0.043, "step": 34700 }, { "epoch": 0.6936173614164102, "grad_norm": 0.1375475823879242, "learning_rate": 9.25730444682664e-05, "loss": 0.0464, "step": 34710 }, { "epoch": 0.6938171935574118, "grad_norm": 0.1484733521938324, "learning_rate": 9.256762412450641e-05, "loss": 0.0395, "step": 34720 }, { "epoch": 0.6940170256984133, "grad_norm": 0.1426737904548645, "learning_rate": 9.256220196233574e-05, "loss": 0.0396, "step": 34730 }, { "epoch": 0.6942168578394149, "grad_norm": 0.10674698650836945, "learning_rate": 9.255677798198596e-05, "loss": 0.0331, "step": 34740 }, { "epoch": 0.6944166899804165, "grad_norm": 0.11590403318405151, "learning_rate": 9.255135218368882e-05, "loss": 0.0322, "step": 34750 }, { "epoch": 0.694616522121418, "grad_norm": 0.1603831797838211, "learning_rate": 9.25459245676761e-05, "loss": 0.0377, "step": 34760 }, { "epoch": 0.6948163542624196, "grad_norm": 0.12928949296474457, "learning_rate": 9.254049513417962e-05, "loss": 0.0368, "step": 34770 }, { "epoch": 0.6950161864034211, "grad_norm": 0.12704338133335114, "learning_rate": 9.253506388343134e-05, "loss": 0.0335, "step": 34780 }, { "epoch": 0.6952160185444227, "grad_norm": 0.14861127734184265, "learning_rate": 9.252963081566328e-05, "loss": 0.0364, "step": 34790 }, { "epoch": 0.6954158506854242, "grad_norm": 0.16417226195335388, "learning_rate": 9.25241959311075e-05, "loss": 0.0345, "step": 34800 }, { "epoch": 0.6956156828264258, "grad_norm": 0.14011333882808685, "learning_rate": 9.251875922999621e-05, "loss": 0.0392, "step": 34810 }, { "epoch": 0.6958155149674273, "grad_norm": 0.10806932300329208, "learning_rate": 9.251332071256164e-05, "loss": 0.0416, "step": 34820 }, { "epoch": 0.6960153471084289, "grad_norm": 0.15047650039196014, "learning_rate": 9.250788037903608e-05, "loss": 0.0404, "step": 34830 }, { "epoch": 0.6962151792494304, "grad_norm": 0.16299161314964294, "learning_rate": 9.250243822965199e-05, "loss": 0.0491, "step": 34840 }, { "epoch": 0.696415011390432, "grad_norm": 0.15477398037910461, "learning_rate": 9.24969942646418e-05, "loss": 0.0436, "step": 34850 }, { "epoch": 0.6966148435314335, "grad_norm": 0.110988549888134, "learning_rate": 9.249154848423808e-05, "loss": 0.04, "step": 34860 }, { "epoch": 0.6968146756724352, "grad_norm": 0.1321258693933487, "learning_rate": 9.248610088867348e-05, "loss": 0.0361, "step": 34870 }, { "epoch": 0.6970145078134368, "grad_norm": 0.20047450065612793, "learning_rate": 9.248065147818071e-05, "loss": 0.0362, "step": 34880 }, { "epoch": 0.6972143399544383, "grad_norm": 0.1511070281267166, "learning_rate": 9.247520025299254e-05, "loss": 0.0388, "step": 34890 }, { "epoch": 0.6974141720954399, "grad_norm": 0.13884377479553223, "learning_rate": 9.246974721334183e-05, "loss": 0.0381, "step": 34900 }, { "epoch": 0.6976140042364414, "grad_norm": 0.1500832587480545, "learning_rate": 9.246429235946155e-05, "loss": 0.0348, "step": 34910 }, { "epoch": 0.697813836377443, "grad_norm": 0.1562904417514801, "learning_rate": 9.245883569158471e-05, "loss": 0.0454, "step": 34920 }, { "epoch": 0.6980136685184445, "grad_norm": 0.2087419629096985, "learning_rate": 9.245337720994441e-05, "loss": 0.0356, "step": 34930 }, { "epoch": 0.6982135006594461, "grad_norm": 0.19806593656539917, "learning_rate": 9.244791691477382e-05, "loss": 0.0446, "step": 34940 }, { "epoch": 0.6984133328004476, "grad_norm": 0.15402594208717346, "learning_rate": 9.24424548063062e-05, "loss": 0.0423, "step": 34950 }, { "epoch": 0.6986131649414492, "grad_norm": 0.12631382048130035, "learning_rate": 9.243699088477489e-05, "loss": 0.0389, "step": 34960 }, { "epoch": 0.6988129970824507, "grad_norm": 0.11816762387752533, "learning_rate": 9.243152515041328e-05, "loss": 0.0389, "step": 34970 }, { "epoch": 0.6990128292234523, "grad_norm": 0.1407492309808731, "learning_rate": 9.242605760345487e-05, "loss": 0.0391, "step": 34980 }, { "epoch": 0.6992126613644538, "grad_norm": 0.11618927866220474, "learning_rate": 9.24205882441332e-05, "loss": 0.0317, "step": 34990 }, { "epoch": 0.6994124935054554, "grad_norm": 0.10012098401784897, "learning_rate": 9.241511707268194e-05, "loss": 0.0306, "step": 35000 }, { "epoch": 0.6996123256464569, "grad_norm": 0.1236242949962616, "learning_rate": 9.240964408933481e-05, "loss": 0.0396, "step": 35010 }, { "epoch": 0.6998121577874585, "grad_norm": 0.1329081505537033, "learning_rate": 9.240416929432558e-05, "loss": 0.0302, "step": 35020 }, { "epoch": 0.7000119899284601, "grad_norm": 0.1138964369893074, "learning_rate": 9.239869268788815e-05, "loss": 0.0444, "step": 35030 }, { "epoch": 0.7002118220694616, "grad_norm": 0.1489546000957489, "learning_rate": 9.239321427025644e-05, "loss": 0.0356, "step": 35040 }, { "epoch": 0.7004116542104633, "grad_norm": 0.12153322249650955, "learning_rate": 9.23877340416645e-05, "loss": 0.0402, "step": 35050 }, { "epoch": 0.7006114863514648, "grad_norm": 0.15484750270843506, "learning_rate": 9.238225200234644e-05, "loss": 0.0465, "step": 35060 }, { "epoch": 0.7008113184924664, "grad_norm": 0.1491706669330597, "learning_rate": 9.237676815253642e-05, "loss": 0.0485, "step": 35070 }, { "epoch": 0.7010111506334679, "grad_norm": 0.1516011357307434, "learning_rate": 9.237128249246871e-05, "loss": 0.0426, "step": 35080 }, { "epoch": 0.7012109827744695, "grad_norm": 0.13804806768894196, "learning_rate": 9.236579502237767e-05, "loss": 0.0493, "step": 35090 }, { "epoch": 0.701410814915471, "grad_norm": 0.15725915133953094, "learning_rate": 9.236030574249767e-05, "loss": 0.0326, "step": 35100 }, { "epoch": 0.7016106470564726, "grad_norm": 0.1635974943637848, "learning_rate": 9.235481465306326e-05, "loss": 0.0356, "step": 35110 }, { "epoch": 0.7018104791974741, "grad_norm": 0.132496640086174, "learning_rate": 9.234932175430895e-05, "loss": 0.0483, "step": 35120 }, { "epoch": 0.7020103113384757, "grad_norm": 0.17583675682544708, "learning_rate": 9.234382704646942e-05, "loss": 0.0373, "step": 35130 }, { "epoch": 0.7022101434794772, "grad_norm": 0.15836893022060394, "learning_rate": 9.23383305297794e-05, "loss": 0.0426, "step": 35140 }, { "epoch": 0.7024099756204788, "grad_norm": 0.20998381078243256, "learning_rate": 9.233283220447366e-05, "loss": 0.0437, "step": 35150 }, { "epoch": 0.7026098077614804, "grad_norm": 0.14108552038669586, "learning_rate": 9.23273320707871e-05, "loss": 0.0466, "step": 35160 }, { "epoch": 0.7028096399024819, "grad_norm": 0.13360491394996643, "learning_rate": 9.232183012895468e-05, "loss": 0.0382, "step": 35170 }, { "epoch": 0.7030094720434835, "grad_norm": 0.1268160045146942, "learning_rate": 9.231632637921142e-05, "loss": 0.0334, "step": 35180 }, { "epoch": 0.703209304184485, "grad_norm": 0.1515268087387085, "learning_rate": 9.231082082179243e-05, "loss": 0.0308, "step": 35190 }, { "epoch": 0.7034091363254866, "grad_norm": 0.10962574183940887, "learning_rate": 9.23053134569329e-05, "loss": 0.0373, "step": 35200 }, { "epoch": 0.7036089684664881, "grad_norm": 0.14680194854736328, "learning_rate": 9.229980428486811e-05, "loss": 0.0309, "step": 35210 }, { "epoch": 0.7038088006074897, "grad_norm": 0.14489105343818665, "learning_rate": 9.229429330583336e-05, "loss": 0.0315, "step": 35220 }, { "epoch": 0.7040086327484912, "grad_norm": 0.13934405148029327, "learning_rate": 9.228878052006413e-05, "loss": 0.0378, "step": 35230 }, { "epoch": 0.7042084648894928, "grad_norm": 0.12461824715137482, "learning_rate": 9.228326592779587e-05, "loss": 0.0389, "step": 35240 }, { "epoch": 0.7044082970304943, "grad_norm": 0.13044363260269165, "learning_rate": 9.227774952926416e-05, "loss": 0.0436, "step": 35250 }, { "epoch": 0.704608129171496, "grad_norm": 0.17309384047985077, "learning_rate": 9.227223132470466e-05, "loss": 0.0354, "step": 35260 }, { "epoch": 0.7048079613124975, "grad_norm": 0.15515607595443726, "learning_rate": 9.226671131435311e-05, "loss": 0.043, "step": 35270 }, { "epoch": 0.7050077934534991, "grad_norm": 0.12980814278125763, "learning_rate": 9.226118949844529e-05, "loss": 0.0436, "step": 35280 }, { "epoch": 0.7052076255945006, "grad_norm": 0.23062461614608765, "learning_rate": 9.225566587721708e-05, "loss": 0.035, "step": 35290 }, { "epoch": 0.7054074577355022, "grad_norm": 0.14549149572849274, "learning_rate": 9.225014045090445e-05, "loss": 0.0384, "step": 35300 }, { "epoch": 0.7056072898765038, "grad_norm": 0.14627981185913086, "learning_rate": 9.224461321974345e-05, "loss": 0.0423, "step": 35310 }, { "epoch": 0.7058071220175053, "grad_norm": 0.15243397653102875, "learning_rate": 9.223908418397017e-05, "loss": 0.0403, "step": 35320 }, { "epoch": 0.7060069541585069, "grad_norm": 0.1554027944803238, "learning_rate": 9.22335533438208e-05, "loss": 0.0497, "step": 35330 }, { "epoch": 0.7062067862995084, "grad_norm": 0.15313608944416046, "learning_rate": 9.222802069953165e-05, "loss": 0.0467, "step": 35340 }, { "epoch": 0.70640661844051, "grad_norm": 0.12545788288116455, "learning_rate": 9.2222486251339e-05, "loss": 0.032, "step": 35350 }, { "epoch": 0.7066064505815115, "grad_norm": 0.14036409556865692, "learning_rate": 9.221694999947931e-05, "loss": 0.0421, "step": 35360 }, { "epoch": 0.7068062827225131, "grad_norm": 0.193065345287323, "learning_rate": 9.221141194418908e-05, "loss": 0.0353, "step": 35370 }, { "epoch": 0.7070061148635146, "grad_norm": 0.14594070613384247, "learning_rate": 9.220587208570485e-05, "loss": 0.0375, "step": 35380 }, { "epoch": 0.7072059470045162, "grad_norm": 0.15603114664554596, "learning_rate": 9.220033042426333e-05, "loss": 0.05, "step": 35390 }, { "epoch": 0.7074057791455177, "grad_norm": 0.16869136691093445, "learning_rate": 9.219478696010122e-05, "loss": 0.0502, "step": 35400 }, { "epoch": 0.7076056112865193, "grad_norm": 0.19781529903411865, "learning_rate": 9.218924169345531e-05, "loss": 0.0425, "step": 35410 }, { "epoch": 0.7078054434275208, "grad_norm": 0.15380482375621796, "learning_rate": 9.21836946245625e-05, "loss": 0.0347, "step": 35420 }, { "epoch": 0.7080052755685224, "grad_norm": 0.1665833741426468, "learning_rate": 9.217814575365975e-05, "loss": 0.0335, "step": 35430 }, { "epoch": 0.708205107709524, "grad_norm": 0.1467149555683136, "learning_rate": 9.21725950809841e-05, "loss": 0.0457, "step": 35440 }, { "epoch": 0.7084049398505256, "grad_norm": 0.12173251062631607, "learning_rate": 9.216704260677264e-05, "loss": 0.0351, "step": 35450 }, { "epoch": 0.7086047719915272, "grad_norm": 0.14844870567321777, "learning_rate": 9.216148833126261e-05, "loss": 0.05, "step": 35460 }, { "epoch": 0.7088046041325287, "grad_norm": 0.13270947337150574, "learning_rate": 9.215593225469125e-05, "loss": 0.0383, "step": 35470 }, { "epoch": 0.7090044362735303, "grad_norm": 0.14826126396656036, "learning_rate": 9.21503743772959e-05, "loss": 0.0388, "step": 35480 }, { "epoch": 0.7092042684145318, "grad_norm": 0.1717289388179779, "learning_rate": 9.2144814699314e-05, "loss": 0.0377, "step": 35490 }, { "epoch": 0.7094041005555334, "grad_norm": 0.12808190286159515, "learning_rate": 9.213925322098302e-05, "loss": 0.0348, "step": 35500 }, { "epoch": 0.7096039326965349, "grad_norm": 0.22138482332229614, "learning_rate": 9.213368994254057e-05, "loss": 0.0344, "step": 35510 }, { "epoch": 0.7098037648375365, "grad_norm": 0.13531170785427094, "learning_rate": 9.212812486422429e-05, "loss": 0.0432, "step": 35520 }, { "epoch": 0.710003596978538, "grad_norm": 0.12439238280057907, "learning_rate": 9.212255798627189e-05, "loss": 0.0484, "step": 35530 }, { "epoch": 0.7102034291195396, "grad_norm": 0.12443576753139496, "learning_rate": 9.211698930892121e-05, "loss": 0.035, "step": 35540 }, { "epoch": 0.7104032612605411, "grad_norm": 0.11662223190069199, "learning_rate": 9.21114188324101e-05, "loss": 0.0321, "step": 35550 }, { "epoch": 0.7106030934015427, "grad_norm": 0.12990593910217285, "learning_rate": 9.210584655697656e-05, "loss": 0.0475, "step": 35560 }, { "epoch": 0.7108029255425443, "grad_norm": 0.12951204180717468, "learning_rate": 9.21002724828586e-05, "loss": 0.0402, "step": 35570 }, { "epoch": 0.7110027576835458, "grad_norm": 0.16917501389980316, "learning_rate": 9.209469661029433e-05, "loss": 0.0337, "step": 35580 }, { "epoch": 0.7112025898245474, "grad_norm": 0.13329729437828064, "learning_rate": 9.208911893952196e-05, "loss": 0.0367, "step": 35590 }, { "epoch": 0.7114024219655489, "grad_norm": 0.1535191535949707, "learning_rate": 9.208353947077974e-05, "loss": 0.0494, "step": 35600 }, { "epoch": 0.7116022541065505, "grad_norm": 0.12314914911985397, "learning_rate": 9.207795820430604e-05, "loss": 0.0411, "step": 35610 }, { "epoch": 0.711802086247552, "grad_norm": 0.12325309216976166, "learning_rate": 9.207237514033925e-05, "loss": 0.0394, "step": 35620 }, { "epoch": 0.7120019183885536, "grad_norm": 0.1296597272157669, "learning_rate": 9.206679027911788e-05, "loss": 0.0329, "step": 35630 }, { "epoch": 0.7122017505295551, "grad_norm": 0.15616704523563385, "learning_rate": 9.206120362088053e-05, "loss": 0.0438, "step": 35640 }, { "epoch": 0.7124015826705568, "grad_norm": 0.16889223456382751, "learning_rate": 9.205561516586581e-05, "loss": 0.0357, "step": 35650 }, { "epoch": 0.7126014148115583, "grad_norm": 0.14589788019657135, "learning_rate": 9.205002491431248e-05, "loss": 0.0412, "step": 35660 }, { "epoch": 0.7128012469525599, "grad_norm": 0.13407568633556366, "learning_rate": 9.204443286645934e-05, "loss": 0.0429, "step": 35670 }, { "epoch": 0.7130010790935614, "grad_norm": 0.18401186168193817, "learning_rate": 9.203883902254525e-05, "loss": 0.0472, "step": 35680 }, { "epoch": 0.713200911234563, "grad_norm": 0.17199663817882538, "learning_rate": 9.20332433828092e-05, "loss": 0.0546, "step": 35690 }, { "epoch": 0.7134007433755645, "grad_norm": 0.1438380777835846, "learning_rate": 9.20276459474902e-05, "loss": 0.0469, "step": 35700 }, { "epoch": 0.7136005755165661, "grad_norm": 0.15409037470817566, "learning_rate": 9.202204671682737e-05, "loss": 0.0418, "step": 35710 }, { "epoch": 0.7138004076575677, "grad_norm": 0.16879430413246155, "learning_rate": 9.20164456910599e-05, "loss": 0.035, "step": 35720 }, { "epoch": 0.7140002397985692, "grad_norm": 0.14343445003032684, "learning_rate": 9.201084287042707e-05, "loss": 0.0392, "step": 35730 }, { "epoch": 0.7142000719395708, "grad_norm": 0.14884068071842194, "learning_rate": 9.20052382551682e-05, "loss": 0.0352, "step": 35740 }, { "epoch": 0.7143999040805723, "grad_norm": 0.14761443436145782, "learning_rate": 9.199963184552272e-05, "loss": 0.0442, "step": 35750 }, { "epoch": 0.7145997362215739, "grad_norm": 0.1472112238407135, "learning_rate": 9.199402364173012e-05, "loss": 0.0437, "step": 35760 }, { "epoch": 0.7147995683625754, "grad_norm": 0.16423480212688446, "learning_rate": 9.198841364402999e-05, "loss": 0.0367, "step": 35770 }, { "epoch": 0.714999400503577, "grad_norm": 0.13872754573822021, "learning_rate": 9.198280185266195e-05, "loss": 0.036, "step": 35780 }, { "epoch": 0.7151992326445785, "grad_norm": 0.14077673852443695, "learning_rate": 9.197718826786574e-05, "loss": 0.0488, "step": 35790 }, { "epoch": 0.7153990647855801, "grad_norm": 0.14187772572040558, "learning_rate": 9.197157288988117e-05, "loss": 0.0343, "step": 35800 }, { "epoch": 0.7155988969265816, "grad_norm": 0.1468491405248642, "learning_rate": 9.196595571894809e-05, "loss": 0.0352, "step": 35810 }, { "epoch": 0.7157987290675832, "grad_norm": 0.13376136124134064, "learning_rate": 9.196033675530648e-05, "loss": 0.0401, "step": 35820 }, { "epoch": 0.7159985612085847, "grad_norm": 0.14424818754196167, "learning_rate": 9.19547159991964e-05, "loss": 0.0367, "step": 35830 }, { "epoch": 0.7161983933495863, "grad_norm": 0.13198421895503998, "learning_rate": 9.194909345085789e-05, "loss": 0.0506, "step": 35840 }, { "epoch": 0.716398225490588, "grad_norm": 0.1291525512933731, "learning_rate": 9.194346911053119e-05, "loss": 0.0456, "step": 35850 }, { "epoch": 0.7165980576315895, "grad_norm": 0.14730378985404968, "learning_rate": 9.193784297845654e-05, "loss": 0.0393, "step": 35860 }, { "epoch": 0.7167978897725911, "grad_norm": 0.19158847630023956, "learning_rate": 9.193221505487427e-05, "loss": 0.0448, "step": 35870 }, { "epoch": 0.7169977219135926, "grad_norm": 0.16264045238494873, "learning_rate": 9.192658534002481e-05, "loss": 0.0419, "step": 35880 }, { "epoch": 0.7171975540545942, "grad_norm": 0.18407060205936432, "learning_rate": 9.192095383414864e-05, "loss": 0.0404, "step": 35890 }, { "epoch": 0.7173973861955957, "grad_norm": 0.15984013676643372, "learning_rate": 9.191532053748635e-05, "loss": 0.0358, "step": 35900 }, { "epoch": 0.7175972183365973, "grad_norm": 0.15082530677318573, "learning_rate": 9.190968545027856e-05, "loss": 0.0456, "step": 35910 }, { "epoch": 0.7177970504775988, "grad_norm": 0.13882815837860107, "learning_rate": 9.1904048572766e-05, "loss": 0.0415, "step": 35920 }, { "epoch": 0.7179968826186004, "grad_norm": 0.18369343876838684, "learning_rate": 9.189840990518946e-05, "loss": 0.0398, "step": 35930 }, { "epoch": 0.7181967147596019, "grad_norm": 0.1401577591896057, "learning_rate": 9.189276944778982e-05, "loss": 0.0461, "step": 35940 }, { "epoch": 0.7183965469006035, "grad_norm": 0.151131272315979, "learning_rate": 9.188712720080802e-05, "loss": 0.0413, "step": 35950 }, { "epoch": 0.718596379041605, "grad_norm": 0.13498015701770782, "learning_rate": 9.18814831644851e-05, "loss": 0.0382, "step": 35960 }, { "epoch": 0.7187962111826066, "grad_norm": 0.15541593730449677, "learning_rate": 9.187583733906218e-05, "loss": 0.0432, "step": 35970 }, { "epoch": 0.7189960433236081, "grad_norm": 0.1215837150812149, "learning_rate": 9.187018972478039e-05, "loss": 0.0394, "step": 35980 }, { "epoch": 0.7191958754646097, "grad_norm": 0.13501906394958496, "learning_rate": 9.186454032188101e-05, "loss": 0.0412, "step": 35990 }, { "epoch": 0.7193957076056113, "grad_norm": 0.1218656525015831, "learning_rate": 9.185888913060539e-05, "loss": 0.0401, "step": 36000 }, { "epoch": 0.7195955397466128, "grad_norm": 0.1369827389717102, "learning_rate": 9.185323615119492e-05, "loss": 0.0448, "step": 36010 }, { "epoch": 0.7197953718876144, "grad_norm": 0.16029122471809387, "learning_rate": 9.184758138389109e-05, "loss": 0.0351, "step": 36020 }, { "epoch": 0.7199952040286159, "grad_norm": 0.16068823635578156, "learning_rate": 9.184192482893547e-05, "loss": 0.0467, "step": 36030 }, { "epoch": 0.7201950361696176, "grad_norm": 0.125371515750885, "learning_rate": 9.183626648656966e-05, "loss": 0.0344, "step": 36040 }, { "epoch": 0.720394868310619, "grad_norm": 0.1488083004951477, "learning_rate": 9.183060635703542e-05, "loss": 0.0449, "step": 36050 }, { "epoch": 0.7205947004516207, "grad_norm": 0.18127967417240143, "learning_rate": 9.182494444057453e-05, "loss": 0.0432, "step": 36060 }, { "epoch": 0.7207945325926222, "grad_norm": 0.1524757295846939, "learning_rate": 9.181928073742884e-05, "loss": 0.0391, "step": 36070 }, { "epoch": 0.7209943647336238, "grad_norm": 0.12335732579231262, "learning_rate": 9.181361524784029e-05, "loss": 0.0399, "step": 36080 }, { "epoch": 0.7211941968746253, "grad_norm": 0.13368330895900726, "learning_rate": 9.180794797205093e-05, "loss": 0.043, "step": 36090 }, { "epoch": 0.7213940290156269, "grad_norm": 0.14325682818889618, "learning_rate": 9.180227891030283e-05, "loss": 0.0409, "step": 36100 }, { "epoch": 0.7215938611566284, "grad_norm": 0.10729849338531494, "learning_rate": 9.179660806283815e-05, "loss": 0.0386, "step": 36110 }, { "epoch": 0.72179369329763, "grad_norm": 0.1568472534418106, "learning_rate": 9.179093542989918e-05, "loss": 0.0409, "step": 36120 }, { "epoch": 0.7219935254386316, "grad_norm": 0.10801646113395691, "learning_rate": 9.178526101172821e-05, "loss": 0.0387, "step": 36130 }, { "epoch": 0.7221933575796331, "grad_norm": 0.12146904319524765, "learning_rate": 9.177958480856765e-05, "loss": 0.0345, "step": 36140 }, { "epoch": 0.7223931897206347, "grad_norm": 0.1763801872730255, "learning_rate": 9.177390682065999e-05, "loss": 0.0369, "step": 36150 }, { "epoch": 0.7225930218616362, "grad_norm": 0.12101996690034866, "learning_rate": 9.176822704824775e-05, "loss": 0.045, "step": 36160 }, { "epoch": 0.7227928540026378, "grad_norm": 0.16423726081848145, "learning_rate": 9.176254549157359e-05, "loss": 0.038, "step": 36170 }, { "epoch": 0.7229926861436393, "grad_norm": 0.1966426819562912, "learning_rate": 9.175686215088022e-05, "loss": 0.0343, "step": 36180 }, { "epoch": 0.7231925182846409, "grad_norm": 0.1450212448835373, "learning_rate": 9.175117702641041e-05, "loss": 0.0394, "step": 36190 }, { "epoch": 0.7233923504256424, "grad_norm": 0.14588147401809692, "learning_rate": 9.1745490118407e-05, "loss": 0.0355, "step": 36200 }, { "epoch": 0.723592182566644, "grad_norm": 0.1591992974281311, "learning_rate": 9.173980142711294e-05, "loss": 0.0445, "step": 36210 }, { "epoch": 0.7237920147076455, "grad_norm": 0.1292058825492859, "learning_rate": 9.173411095277125e-05, "loss": 0.04, "step": 36220 }, { "epoch": 0.7239918468486471, "grad_norm": 0.1275026798248291, "learning_rate": 9.1728418695625e-05, "loss": 0.0346, "step": 36230 }, { "epoch": 0.7241916789896486, "grad_norm": 0.13260336220264435, "learning_rate": 9.172272465591739e-05, "loss": 0.0473, "step": 36240 }, { "epoch": 0.7243915111306503, "grad_norm": 0.10946843028068542, "learning_rate": 9.171702883389159e-05, "loss": 0.0299, "step": 36250 }, { "epoch": 0.7245913432716519, "grad_norm": 0.15471267700195312, "learning_rate": 9.171133122979098e-05, "loss": 0.0379, "step": 36260 }, { "epoch": 0.7247911754126534, "grad_norm": 0.16083891689777374, "learning_rate": 9.170563184385892e-05, "loss": 0.042, "step": 36270 }, { "epoch": 0.724991007553655, "grad_norm": 0.13815738260746002, "learning_rate": 9.169993067633888e-05, "loss": 0.0484, "step": 36280 }, { "epoch": 0.7251908396946565, "grad_norm": 0.12586449086666107, "learning_rate": 9.16942277274744e-05, "loss": 0.0374, "step": 36290 }, { "epoch": 0.7253906718356581, "grad_norm": 0.14731666445732117, "learning_rate": 9.168852299750912e-05, "loss": 0.0317, "step": 36300 }, { "epoch": 0.7255905039766596, "grad_norm": 0.1323060244321823, "learning_rate": 9.168281648668671e-05, "loss": 0.039, "step": 36310 }, { "epoch": 0.7257903361176612, "grad_norm": 0.11620424687862396, "learning_rate": 9.167710819525097e-05, "loss": 0.0469, "step": 36320 }, { "epoch": 0.7259901682586627, "grad_norm": 0.14698746800422668, "learning_rate": 9.167139812344572e-05, "loss": 0.0272, "step": 36330 }, { "epoch": 0.7261900003996643, "grad_norm": 0.1245177611708641, "learning_rate": 9.16656862715149e-05, "loss": 0.0428, "step": 36340 }, { "epoch": 0.7263898325406658, "grad_norm": 0.12132585048675537, "learning_rate": 9.16599726397025e-05, "loss": 0.0366, "step": 36350 }, { "epoch": 0.7265896646816674, "grad_norm": 0.13166455924510956, "learning_rate": 9.16542572282526e-05, "loss": 0.0461, "step": 36360 }, { "epoch": 0.7267894968226689, "grad_norm": 0.1983836591243744, "learning_rate": 9.164854003740936e-05, "loss": 0.0461, "step": 36370 }, { "epoch": 0.7269893289636705, "grad_norm": 0.15626220405101776, "learning_rate": 9.164282106741699e-05, "loss": 0.0405, "step": 36380 }, { "epoch": 0.727189161104672, "grad_norm": 0.14907126128673553, "learning_rate": 9.163710031851983e-05, "loss": 0.0323, "step": 36390 }, { "epoch": 0.7273889932456736, "grad_norm": 0.13162803649902344, "learning_rate": 9.16313777909622e-05, "loss": 0.035, "step": 36400 }, { "epoch": 0.7275888253866752, "grad_norm": 0.12353487312793732, "learning_rate": 9.162565348498859e-05, "loss": 0.029, "step": 36410 }, { "epoch": 0.7277886575276767, "grad_norm": 0.1221974566578865, "learning_rate": 9.161992740084356e-05, "loss": 0.0347, "step": 36420 }, { "epoch": 0.7279884896686784, "grad_norm": 0.18262305855751038, "learning_rate": 9.161419953877168e-05, "loss": 0.0351, "step": 36430 }, { "epoch": 0.7281883218096799, "grad_norm": 0.15061098337173462, "learning_rate": 9.160846989901764e-05, "loss": 0.0429, "step": 36440 }, { "epoch": 0.7283881539506815, "grad_norm": 0.14482101798057556, "learning_rate": 9.16027384818262e-05, "loss": 0.0341, "step": 36450 }, { "epoch": 0.728587986091683, "grad_norm": 0.11254958063364029, "learning_rate": 9.15970052874422e-05, "loss": 0.0448, "step": 36460 }, { "epoch": 0.7287878182326846, "grad_norm": 0.17353741824626923, "learning_rate": 9.159127031611054e-05, "loss": 0.0426, "step": 36470 }, { "epoch": 0.7289876503736861, "grad_norm": 0.1557108759880066, "learning_rate": 9.158553356807624e-05, "loss": 0.0405, "step": 36480 }, { "epoch": 0.7291874825146877, "grad_norm": 0.15121163427829742, "learning_rate": 9.157979504358431e-05, "loss": 0.042, "step": 36490 }, { "epoch": 0.7293873146556892, "grad_norm": 0.2056901752948761, "learning_rate": 9.157405474287995e-05, "loss": 0.047, "step": 36500 }, { "epoch": 0.7295871467966908, "grad_norm": 0.15757356584072113, "learning_rate": 9.156831266620833e-05, "loss": 0.0388, "step": 36510 }, { "epoch": 0.7297869789376923, "grad_norm": 0.12295841425657272, "learning_rate": 9.156256881381476e-05, "loss": 0.0303, "step": 36520 }, { "epoch": 0.7299868110786939, "grad_norm": 0.12223672866821289, "learning_rate": 9.155682318594461e-05, "loss": 0.0424, "step": 36530 }, { "epoch": 0.7301866432196955, "grad_norm": 0.1255662441253662, "learning_rate": 9.15510757828433e-05, "loss": 0.0349, "step": 36540 }, { "epoch": 0.730386475360697, "grad_norm": 0.12238797545433044, "learning_rate": 9.154532660475639e-05, "loss": 0.042, "step": 36550 }, { "epoch": 0.7305863075016986, "grad_norm": 0.11088468879461288, "learning_rate": 9.153957565192944e-05, "loss": 0.0259, "step": 36560 }, { "epoch": 0.7307861396427001, "grad_norm": 0.1676563024520874, "learning_rate": 9.15338229246081e-05, "loss": 0.0436, "step": 36570 }, { "epoch": 0.7309859717837017, "grad_norm": 0.1614043116569519, "learning_rate": 9.152806842303819e-05, "loss": 0.0425, "step": 36580 }, { "epoch": 0.7311858039247032, "grad_norm": 0.10456538200378418, "learning_rate": 9.152231214746545e-05, "loss": 0.044, "step": 36590 }, { "epoch": 0.7313856360657048, "grad_norm": 0.1601618081331253, "learning_rate": 9.151655409813582e-05, "loss": 0.0434, "step": 36600 }, { "epoch": 0.7315854682067063, "grad_norm": 0.14967100322246552, "learning_rate": 9.151079427529527e-05, "loss": 0.038, "step": 36610 }, { "epoch": 0.731785300347708, "grad_norm": 0.1628722846508026, "learning_rate": 9.150503267918987e-05, "loss": 0.0378, "step": 36620 }, { "epoch": 0.7319851324887094, "grad_norm": 0.11722591519355774, "learning_rate": 9.149926931006569e-05, "loss": 0.0318, "step": 36630 }, { "epoch": 0.732184964629711, "grad_norm": 0.10498633980751038, "learning_rate": 9.149350416816898e-05, "loss": 0.0334, "step": 36640 }, { "epoch": 0.7323847967707126, "grad_norm": 0.14389443397521973, "learning_rate": 9.1487737253746e-05, "loss": 0.0399, "step": 36650 }, { "epoch": 0.7325846289117142, "grad_norm": 0.1266937106847763, "learning_rate": 9.148196856704307e-05, "loss": 0.0384, "step": 36660 }, { "epoch": 0.7327844610527158, "grad_norm": 0.14141885936260223, "learning_rate": 9.147619810830667e-05, "loss": 0.0342, "step": 36670 }, { "epoch": 0.7329842931937173, "grad_norm": 0.1189345270395279, "learning_rate": 9.147042587778329e-05, "loss": 0.0362, "step": 36680 }, { "epoch": 0.7331841253347189, "grad_norm": 0.22609324753284454, "learning_rate": 9.146465187571949e-05, "loss": 0.039, "step": 36690 }, { "epoch": 0.7333839574757204, "grad_norm": 0.1376475691795349, "learning_rate": 9.145887610236194e-05, "loss": 0.0298, "step": 36700 }, { "epoch": 0.733583789616722, "grad_norm": 0.1170196458697319, "learning_rate": 9.145309855795736e-05, "loss": 0.0408, "step": 36710 }, { "epoch": 0.7337836217577235, "grad_norm": 0.10890171676874161, "learning_rate": 9.144731924275256e-05, "loss": 0.0417, "step": 36720 }, { "epoch": 0.7339834538987251, "grad_norm": 0.12496199458837509, "learning_rate": 9.144153815699443e-05, "loss": 0.0399, "step": 36730 }, { "epoch": 0.7341832860397266, "grad_norm": 0.14547081291675568, "learning_rate": 9.143575530092992e-05, "loss": 0.0491, "step": 36740 }, { "epoch": 0.7343831181807282, "grad_norm": 0.1219387948513031, "learning_rate": 9.142997067480606e-05, "loss": 0.0365, "step": 36750 }, { "epoch": 0.7345829503217297, "grad_norm": 0.1518462598323822, "learning_rate": 9.142418427886995e-05, "loss": 0.0458, "step": 36760 }, { "epoch": 0.7347827824627313, "grad_norm": 0.1521470695734024, "learning_rate": 9.14183961133688e-05, "loss": 0.0374, "step": 36770 }, { "epoch": 0.7349826146037328, "grad_norm": 0.16887296736240387, "learning_rate": 9.141260617854986e-05, "loss": 0.0379, "step": 36780 }, { "epoch": 0.7351824467447344, "grad_norm": 0.1505117267370224, "learning_rate": 9.140681447466044e-05, "loss": 0.0523, "step": 36790 }, { "epoch": 0.7353822788857359, "grad_norm": 0.1633392870426178, "learning_rate": 9.1401021001948e-05, "loss": 0.0337, "step": 36800 }, { "epoch": 0.7355821110267375, "grad_norm": 0.12528514862060547, "learning_rate": 9.139522576065998e-05, "loss": 0.0415, "step": 36810 }, { "epoch": 0.7357819431677391, "grad_norm": 0.15078330039978027, "learning_rate": 9.138942875104397e-05, "loss": 0.0433, "step": 36820 }, { "epoch": 0.7359817753087406, "grad_norm": 0.14351913332939148, "learning_rate": 9.138362997334759e-05, "loss": 0.0436, "step": 36830 }, { "epoch": 0.7361816074497423, "grad_norm": 0.15464989840984344, "learning_rate": 9.137782942781856e-05, "loss": 0.0357, "step": 36840 }, { "epoch": 0.7363814395907438, "grad_norm": 0.1917605698108673, "learning_rate": 9.137202711470469e-05, "loss": 0.0403, "step": 36850 }, { "epoch": 0.7365812717317454, "grad_norm": 0.1736113280057907, "learning_rate": 9.13662230342538e-05, "loss": 0.0405, "step": 36860 }, { "epoch": 0.7367811038727469, "grad_norm": 0.14856071770191193, "learning_rate": 9.136041718671388e-05, "loss": 0.0369, "step": 36870 }, { "epoch": 0.7369809360137485, "grad_norm": 0.13392922282218933, "learning_rate": 9.135460957233291e-05, "loss": 0.0411, "step": 36880 }, { "epoch": 0.73718076815475, "grad_norm": 0.16179466247558594, "learning_rate": 9.134880019135897e-05, "loss": 0.0401, "step": 36890 }, { "epoch": 0.7373806002957516, "grad_norm": 0.168531134724617, "learning_rate": 9.134298904404026e-05, "loss": 0.0377, "step": 36900 }, { "epoch": 0.7375804324367531, "grad_norm": 0.16803355515003204, "learning_rate": 9.133717613062503e-05, "loss": 0.0418, "step": 36910 }, { "epoch": 0.7377802645777547, "grad_norm": 0.1543780416250229, "learning_rate": 9.133136145136154e-05, "loss": 0.0414, "step": 36920 }, { "epoch": 0.7379800967187562, "grad_norm": 0.1556757688522339, "learning_rate": 9.132554500649823e-05, "loss": 0.0384, "step": 36930 }, { "epoch": 0.7381799288597578, "grad_norm": 0.12177512794733047, "learning_rate": 9.131972679628355e-05, "loss": 0.0352, "step": 36940 }, { "epoch": 0.7383797610007594, "grad_norm": 0.13884411752223969, "learning_rate": 9.131390682096605e-05, "loss": 0.0309, "step": 36950 }, { "epoch": 0.7385795931417609, "grad_norm": 0.15317270159721375, "learning_rate": 9.130808508079435e-05, "loss": 0.0395, "step": 36960 }, { "epoch": 0.7387794252827625, "grad_norm": 0.14500105381011963, "learning_rate": 9.130226157601714e-05, "loss": 0.0454, "step": 36970 }, { "epoch": 0.738979257423764, "grad_norm": 0.17499157786369324, "learning_rate": 9.12964363068832e-05, "loss": 0.0363, "step": 36980 }, { "epoch": 0.7391790895647656, "grad_norm": 0.15874509513378143, "learning_rate": 9.129060927364135e-05, "loss": 0.04, "step": 36990 }, { "epoch": 0.7393789217057671, "grad_norm": 0.25787338614463806, "learning_rate": 9.128478047654055e-05, "loss": 0.0373, "step": 37000 }, { "epoch": 0.7395787538467687, "grad_norm": 0.18411476910114288, "learning_rate": 9.127894991582974e-05, "loss": 0.043, "step": 37010 }, { "epoch": 0.7397785859877702, "grad_norm": 0.13628743588924408, "learning_rate": 9.127311759175803e-05, "loss": 0.0419, "step": 37020 }, { "epoch": 0.7399784181287719, "grad_norm": 0.14296433329582214, "learning_rate": 9.126728350457456e-05, "loss": 0.0394, "step": 37030 }, { "epoch": 0.7401782502697734, "grad_norm": 0.15253645181655884, "learning_rate": 9.126144765452855e-05, "loss": 0.0353, "step": 37040 }, { "epoch": 0.740378082410775, "grad_norm": 0.15495353937149048, "learning_rate": 9.12556100418693e-05, "loss": 0.0359, "step": 37050 }, { "epoch": 0.7405779145517765, "grad_norm": 0.17527367174625397, "learning_rate": 9.124977066684617e-05, "loss": 0.0459, "step": 37060 }, { "epoch": 0.7407777466927781, "grad_norm": 0.13331492245197296, "learning_rate": 9.124392952970861e-05, "loss": 0.0486, "step": 37070 }, { "epoch": 0.7409775788337796, "grad_norm": 0.12737970054149628, "learning_rate": 9.123808663070617e-05, "loss": 0.0471, "step": 37080 }, { "epoch": 0.7411774109747812, "grad_norm": 0.1668684184551239, "learning_rate": 9.123224197008841e-05, "loss": 0.0424, "step": 37090 }, { "epoch": 0.7413772431157828, "grad_norm": 0.10726424306631088, "learning_rate": 9.122639554810501e-05, "loss": 0.0308, "step": 37100 }, { "epoch": 0.7415770752567843, "grad_norm": 0.16066743433475494, "learning_rate": 9.122054736500576e-05, "loss": 0.0431, "step": 37110 }, { "epoch": 0.7417769073977859, "grad_norm": 0.18442559242248535, "learning_rate": 9.121469742104043e-05, "loss": 0.048, "step": 37120 }, { "epoch": 0.7419767395387874, "grad_norm": 0.1351742148399353, "learning_rate": 9.120884571645894e-05, "loss": 0.0373, "step": 37130 }, { "epoch": 0.742176571679789, "grad_norm": 0.12385613471269608, "learning_rate": 9.120299225151126e-05, "loss": 0.0416, "step": 37140 }, { "epoch": 0.7423764038207905, "grad_norm": 0.13921475410461426, "learning_rate": 9.119713702644744e-05, "loss": 0.026, "step": 37150 }, { "epoch": 0.7425762359617921, "grad_norm": 0.14943307638168335, "learning_rate": 9.119128004151762e-05, "loss": 0.0476, "step": 37160 }, { "epoch": 0.7427760681027936, "grad_norm": 0.1297353059053421, "learning_rate": 9.118542129697199e-05, "loss": 0.0378, "step": 37170 }, { "epoch": 0.7429759002437952, "grad_norm": 0.1731397658586502, "learning_rate": 9.117956079306081e-05, "loss": 0.0474, "step": 37180 }, { "epoch": 0.7431757323847967, "grad_norm": 0.1429404765367508, "learning_rate": 9.117369853003445e-05, "loss": 0.0434, "step": 37190 }, { "epoch": 0.7433755645257983, "grad_norm": 0.16009113192558289, "learning_rate": 9.116783450814332e-05, "loss": 0.0442, "step": 37200 }, { "epoch": 0.7435753966667998, "grad_norm": 0.13449294865131378, "learning_rate": 9.116196872763793e-05, "loss": 0.0429, "step": 37210 }, { "epoch": 0.7437752288078014, "grad_norm": 0.14202436804771423, "learning_rate": 9.115610118876886e-05, "loss": 0.04, "step": 37220 }, { "epoch": 0.7439750609488031, "grad_norm": 0.12814486026763916, "learning_rate": 9.115023189178676e-05, "loss": 0.0329, "step": 37230 }, { "epoch": 0.7441748930898046, "grad_norm": 0.1490713506937027, "learning_rate": 9.114436083694234e-05, "loss": 0.0361, "step": 37240 }, { "epoch": 0.7443747252308062, "grad_norm": 0.18741345405578613, "learning_rate": 9.113848802448639e-05, "loss": 0.0354, "step": 37250 }, { "epoch": 0.7445745573718077, "grad_norm": 0.1336815059185028, "learning_rate": 9.113261345466983e-05, "loss": 0.0388, "step": 37260 }, { "epoch": 0.7447743895128093, "grad_norm": 0.17809905111789703, "learning_rate": 9.112673712774358e-05, "loss": 0.0331, "step": 37270 }, { "epoch": 0.7449742216538108, "grad_norm": 0.12224772572517395, "learning_rate": 9.112085904395866e-05, "loss": 0.035, "step": 37280 }, { "epoch": 0.7451740537948124, "grad_norm": 0.12790930271148682, "learning_rate": 9.11149792035662e-05, "loss": 0.039, "step": 37290 }, { "epoch": 0.7453738859358139, "grad_norm": 0.1460992693901062, "learning_rate": 9.110909760681736e-05, "loss": 0.0382, "step": 37300 }, { "epoch": 0.7455737180768155, "grad_norm": 0.1270104944705963, "learning_rate": 9.110321425396338e-05, "loss": 0.037, "step": 37310 }, { "epoch": 0.745773550217817, "grad_norm": 0.1574813723564148, "learning_rate": 9.109732914525559e-05, "loss": 0.0401, "step": 37320 }, { "epoch": 0.7459733823588186, "grad_norm": 0.1401900053024292, "learning_rate": 9.109144228094541e-05, "loss": 0.0494, "step": 37330 }, { "epoch": 0.7461732144998201, "grad_norm": 0.19639036059379578, "learning_rate": 9.108555366128431e-05, "loss": 0.05, "step": 37340 }, { "epoch": 0.7463730466408217, "grad_norm": 0.13810212910175323, "learning_rate": 9.107966328652382e-05, "loss": 0.0415, "step": 37350 }, { "epoch": 0.7465728787818233, "grad_norm": 0.12537728250026703, "learning_rate": 9.107377115691557e-05, "loss": 0.0333, "step": 37360 }, { "epoch": 0.7467727109228248, "grad_norm": 0.12826840579509735, "learning_rate": 9.106787727271129e-05, "loss": 0.0428, "step": 37370 }, { "epoch": 0.7469725430638264, "grad_norm": 0.15672771632671356, "learning_rate": 9.106198163416272e-05, "loss": 0.0424, "step": 37380 }, { "epoch": 0.7471723752048279, "grad_norm": 0.17026016116142273, "learning_rate": 9.105608424152176e-05, "loss": 0.0385, "step": 37390 }, { "epoch": 0.7473722073458295, "grad_norm": 0.16365638375282288, "learning_rate": 9.105018509504029e-05, "loss": 0.0412, "step": 37400 }, { "epoch": 0.747572039486831, "grad_norm": 0.13396354019641876, "learning_rate": 9.104428419497031e-05, "loss": 0.044, "step": 37410 }, { "epoch": 0.7477718716278327, "grad_norm": 0.14237673580646515, "learning_rate": 9.103838154156392e-05, "loss": 0.036, "step": 37420 }, { "epoch": 0.7479717037688342, "grad_norm": 0.12948143482208252, "learning_rate": 9.103247713507325e-05, "loss": 0.0402, "step": 37430 }, { "epoch": 0.7481715359098358, "grad_norm": 0.18355214595794678, "learning_rate": 9.102657097575056e-05, "loss": 0.0331, "step": 37440 }, { "epoch": 0.7483713680508373, "grad_norm": 0.16184477508068085, "learning_rate": 9.10206630638481e-05, "loss": 0.0344, "step": 37450 }, { "epoch": 0.7485712001918389, "grad_norm": 0.1835218220949173, "learning_rate": 9.101475339961829e-05, "loss": 0.0378, "step": 37460 }, { "epoch": 0.7487710323328404, "grad_norm": 0.14823192358016968, "learning_rate": 9.100884198331354e-05, "loss": 0.0424, "step": 37470 }, { "epoch": 0.748970864473842, "grad_norm": 0.15988598763942719, "learning_rate": 9.100292881518642e-05, "loss": 0.0352, "step": 37480 }, { "epoch": 0.7491706966148435, "grad_norm": 0.1772805005311966, "learning_rate": 9.099701389548949e-05, "loss": 0.0335, "step": 37490 }, { "epoch": 0.7493705287558451, "grad_norm": 0.15231801569461823, "learning_rate": 9.099109722447546e-05, "loss": 0.0342, "step": 37500 }, { "epoch": 0.7495703608968467, "grad_norm": 0.11302407085895538, "learning_rate": 9.098517880239703e-05, "loss": 0.0372, "step": 37510 }, { "epoch": 0.7497701930378482, "grad_norm": 0.1184363141655922, "learning_rate": 9.097925862950706e-05, "loss": 0.0363, "step": 37520 }, { "epoch": 0.7499700251788498, "grad_norm": 0.1585322767496109, "learning_rate": 9.097333670605846e-05, "loss": 0.0398, "step": 37530 }, { "epoch": 0.7501698573198513, "grad_norm": 0.13108402490615845, "learning_rate": 9.096741303230418e-05, "loss": 0.04, "step": 37540 }, { "epoch": 0.7503696894608529, "grad_norm": 0.15639400482177734, "learning_rate": 9.096148760849728e-05, "loss": 0.0363, "step": 37550 }, { "epoch": 0.7505695216018544, "grad_norm": 0.17556244134902954, "learning_rate": 9.095556043489086e-05, "loss": 0.0395, "step": 37560 }, { "epoch": 0.750769353742856, "grad_norm": 0.11561179161071777, "learning_rate": 9.094963151173814e-05, "loss": 0.0397, "step": 37570 }, { "epoch": 0.7509691858838575, "grad_norm": 0.1374957263469696, "learning_rate": 9.09437008392924e-05, "loss": 0.0406, "step": 37580 }, { "epoch": 0.7511690180248591, "grad_norm": 0.13549523055553436, "learning_rate": 9.093776841780696e-05, "loss": 0.0407, "step": 37590 }, { "epoch": 0.7513688501658606, "grad_norm": 0.15455424785614014, "learning_rate": 9.093183424753528e-05, "loss": 0.0336, "step": 37600 }, { "epoch": 0.7515686823068622, "grad_norm": 0.1493762880563736, "learning_rate": 9.092589832873082e-05, "loss": 0.0494, "step": 37610 }, { "epoch": 0.7517685144478637, "grad_norm": 0.11374098807573318, "learning_rate": 9.091996066164717e-05, "loss": 0.0371, "step": 37620 }, { "epoch": 0.7519683465888654, "grad_norm": 0.12294380366802216, "learning_rate": 9.091402124653798e-05, "loss": 0.0347, "step": 37630 }, { "epoch": 0.752168178729867, "grad_norm": 0.12523260712623596, "learning_rate": 9.090808008365697e-05, "loss": 0.0445, "step": 37640 }, { "epoch": 0.7523680108708685, "grad_norm": 0.18138107657432556, "learning_rate": 9.09021371732579e-05, "loss": 0.039, "step": 37650 }, { "epoch": 0.7525678430118701, "grad_norm": 0.1393466740846634, "learning_rate": 9.089619251559468e-05, "loss": 0.0376, "step": 37660 }, { "epoch": 0.7527676751528716, "grad_norm": 0.14667873084545135, "learning_rate": 9.089024611092127e-05, "loss": 0.0362, "step": 37670 }, { "epoch": 0.7529675072938732, "grad_norm": 0.13486164808273315, "learning_rate": 9.088429795949164e-05, "loss": 0.0334, "step": 37680 }, { "epoch": 0.7531673394348747, "grad_norm": 0.1577693074941635, "learning_rate": 9.08783480615599e-05, "loss": 0.0438, "step": 37690 }, { "epoch": 0.7533671715758763, "grad_norm": 0.1453433334827423, "learning_rate": 9.087239641738025e-05, "loss": 0.0383, "step": 37700 }, { "epoch": 0.7535670037168778, "grad_norm": 0.14165127277374268, "learning_rate": 9.08664430272069e-05, "loss": 0.0347, "step": 37710 }, { "epoch": 0.7537668358578794, "grad_norm": 0.12711212038993835, "learning_rate": 9.086048789129417e-05, "loss": 0.0389, "step": 37720 }, { "epoch": 0.7539666679988809, "grad_norm": 0.12218952924013138, "learning_rate": 9.085453100989648e-05, "loss": 0.0352, "step": 37730 }, { "epoch": 0.7541665001398825, "grad_norm": 0.1299598067998886, "learning_rate": 9.084857238326827e-05, "loss": 0.0413, "step": 37740 }, { "epoch": 0.754366332280884, "grad_norm": 0.11841169744729996, "learning_rate": 9.084261201166408e-05, "loss": 0.0355, "step": 37750 }, { "epoch": 0.7545661644218856, "grad_norm": 0.14596877992153168, "learning_rate": 9.083664989533855e-05, "loss": 0.0454, "step": 37760 }, { "epoch": 0.7547659965628871, "grad_norm": 0.13829012215137482, "learning_rate": 9.083068603454635e-05, "loss": 0.0319, "step": 37770 }, { "epoch": 0.7549658287038887, "grad_norm": 0.10632368177175522, "learning_rate": 9.082472042954223e-05, "loss": 0.038, "step": 37780 }, { "epoch": 0.7551656608448903, "grad_norm": 0.20117338001728058, "learning_rate": 9.081875308058107e-05, "loss": 0.0335, "step": 37790 }, { "epoch": 0.7553654929858918, "grad_norm": 0.13410469889640808, "learning_rate": 9.081278398791777e-05, "loss": 0.0442, "step": 37800 }, { "epoch": 0.7555653251268935, "grad_norm": 0.15077899396419525, "learning_rate": 9.08068131518073e-05, "loss": 0.0359, "step": 37810 }, { "epoch": 0.755765157267895, "grad_norm": 0.14202268421649933, "learning_rate": 9.080084057250475e-05, "loss": 0.0359, "step": 37820 }, { "epoch": 0.7559649894088966, "grad_norm": 0.1571885496377945, "learning_rate": 9.079486625026523e-05, "loss": 0.047, "step": 37830 }, { "epoch": 0.7561648215498981, "grad_norm": 0.1667308211326599, "learning_rate": 9.078889018534398e-05, "loss": 0.0459, "step": 37840 }, { "epoch": 0.7563646536908997, "grad_norm": 0.16347435116767883, "learning_rate": 9.078291237799626e-05, "loss": 0.0455, "step": 37850 }, { "epoch": 0.7565644858319012, "grad_norm": 0.11866562813520432, "learning_rate": 9.077693282847744e-05, "loss": 0.0327, "step": 37860 }, { "epoch": 0.7567643179729028, "grad_norm": 0.14738944172859192, "learning_rate": 9.077095153704298e-05, "loss": 0.0411, "step": 37870 }, { "epoch": 0.7569641501139043, "grad_norm": 0.15398406982421875, "learning_rate": 9.076496850394836e-05, "loss": 0.0379, "step": 37880 }, { "epoch": 0.7571639822549059, "grad_norm": 0.11416569352149963, "learning_rate": 9.075898372944916e-05, "loss": 0.0349, "step": 37890 }, { "epoch": 0.7573638143959074, "grad_norm": 0.12761163711547852, "learning_rate": 9.075299721380107e-05, "loss": 0.0337, "step": 37900 }, { "epoch": 0.757563646536909, "grad_norm": 0.15336517989635468, "learning_rate": 9.07470089572598e-05, "loss": 0.044, "step": 37910 }, { "epoch": 0.7577634786779106, "grad_norm": 0.16851255297660828, "learning_rate": 9.074101896008116e-05, "loss": 0.0423, "step": 37920 }, { "epoch": 0.7579633108189121, "grad_norm": 0.17745716869831085, "learning_rate": 9.073502722252104e-05, "loss": 0.0427, "step": 37930 }, { "epoch": 0.7581631429599137, "grad_norm": 0.1328703612089157, "learning_rate": 9.072903374483537e-05, "loss": 0.0306, "step": 37940 }, { "epoch": 0.7583629751009152, "grad_norm": 0.12691529095172882, "learning_rate": 9.072303852728021e-05, "loss": 0.0384, "step": 37950 }, { "epoch": 0.7585628072419168, "grad_norm": 0.13479670882225037, "learning_rate": 9.071704157011166e-05, "loss": 0.0347, "step": 37960 }, { "epoch": 0.7587626393829183, "grad_norm": 0.1251230239868164, "learning_rate": 9.07110428735859e-05, "loss": 0.0326, "step": 37970 }, { "epoch": 0.7589624715239199, "grad_norm": 0.13248498737812042, "learning_rate": 9.070504243795916e-05, "loss": 0.031, "step": 37980 }, { "epoch": 0.7591623036649214, "grad_norm": 0.23243917524814606, "learning_rate": 9.06990402634878e-05, "loss": 0.0362, "step": 37990 }, { "epoch": 0.759362135805923, "grad_norm": 0.14198879897594452, "learning_rate": 9.06930363504282e-05, "loss": 0.0306, "step": 38000 }, { "epoch": 0.7595619679469245, "grad_norm": 0.16492509841918945, "learning_rate": 9.068703069903684e-05, "loss": 0.0406, "step": 38010 }, { "epoch": 0.7597618000879262, "grad_norm": 0.1557471752166748, "learning_rate": 9.068102330957027e-05, "loss": 0.036, "step": 38020 }, { "epoch": 0.7599616322289277, "grad_norm": 0.15797406435012817, "learning_rate": 9.067501418228513e-05, "loss": 0.0467, "step": 38030 }, { "epoch": 0.7601614643699293, "grad_norm": 0.16882354021072388, "learning_rate": 9.066900331743809e-05, "loss": 0.0384, "step": 38040 }, { "epoch": 0.7603612965109309, "grad_norm": 0.1418028026819229, "learning_rate": 9.066299071528594e-05, "loss": 0.0438, "step": 38050 }, { "epoch": 0.7605611286519324, "grad_norm": 0.15405942499637604, "learning_rate": 9.065697637608554e-05, "loss": 0.0428, "step": 38060 }, { "epoch": 0.760760960792934, "grad_norm": 0.150117427110672, "learning_rate": 9.06509603000938e-05, "loss": 0.0369, "step": 38070 }, { "epoch": 0.7609607929339355, "grad_norm": 0.1346570998430252, "learning_rate": 9.064494248756771e-05, "loss": 0.0424, "step": 38080 }, { "epoch": 0.7611606250749371, "grad_norm": 0.1804920881986618, "learning_rate": 9.063892293876434e-05, "loss": 0.0381, "step": 38090 }, { "epoch": 0.7613604572159386, "grad_norm": 0.16875648498535156, "learning_rate": 9.063290165394085e-05, "loss": 0.0326, "step": 38100 }, { "epoch": 0.7615602893569402, "grad_norm": 0.1164594367146492, "learning_rate": 9.062687863335442e-05, "loss": 0.0299, "step": 38110 }, { "epoch": 0.7617601214979417, "grad_norm": 0.13788647949695587, "learning_rate": 9.06208538772624e-05, "loss": 0.0389, "step": 38120 }, { "epoch": 0.7619599536389433, "grad_norm": 0.2094956636428833, "learning_rate": 9.06148273859221e-05, "loss": 0.0387, "step": 38130 }, { "epoch": 0.7621597857799448, "grad_norm": 0.13235852122306824, "learning_rate": 9.0608799159591e-05, "loss": 0.0392, "step": 38140 }, { "epoch": 0.7623596179209464, "grad_norm": 0.15663988888263702, "learning_rate": 9.06027691985266e-05, "loss": 0.0368, "step": 38150 }, { "epoch": 0.7625594500619479, "grad_norm": 0.1748238503932953, "learning_rate": 9.05967375029865e-05, "loss": 0.0415, "step": 38160 }, { "epoch": 0.7627592822029495, "grad_norm": 0.1781635880470276, "learning_rate": 9.059070407322833e-05, "loss": 0.0319, "step": 38170 }, { "epoch": 0.762959114343951, "grad_norm": 0.14127226173877716, "learning_rate": 9.058466890950985e-05, "loss": 0.0425, "step": 38180 }, { "epoch": 0.7631589464849526, "grad_norm": 0.18054300546646118, "learning_rate": 9.057863201208888e-05, "loss": 0.039, "step": 38190 }, { "epoch": 0.7633587786259542, "grad_norm": 0.13938814401626587, "learning_rate": 9.057259338122331e-05, "loss": 0.0409, "step": 38200 }, { "epoch": 0.7635586107669557, "grad_norm": 0.1507779061794281, "learning_rate": 9.056655301717106e-05, "loss": 0.0372, "step": 38210 }, { "epoch": 0.7637584429079574, "grad_norm": 0.12930001318454742, "learning_rate": 9.05605109201902e-05, "loss": 0.035, "step": 38220 }, { "epoch": 0.7639582750489589, "grad_norm": 0.13984942436218262, "learning_rate": 9.055446709053883e-05, "loss": 0.0409, "step": 38230 }, { "epoch": 0.7641581071899605, "grad_norm": 0.14067508280277252, "learning_rate": 9.054842152847512e-05, "loss": 0.0294, "step": 38240 }, { "epoch": 0.764357939330962, "grad_norm": 0.11903391778469086, "learning_rate": 9.054237423425733e-05, "loss": 0.0378, "step": 38250 }, { "epoch": 0.7645577714719636, "grad_norm": 0.12671710550785065, "learning_rate": 9.05363252081438e-05, "loss": 0.0422, "step": 38260 }, { "epoch": 0.7647576036129651, "grad_norm": 0.1272997260093689, "learning_rate": 9.05302744503929e-05, "loss": 0.039, "step": 38270 }, { "epoch": 0.7649574357539667, "grad_norm": 0.12400894612073898, "learning_rate": 9.052422196126316e-05, "loss": 0.0327, "step": 38280 }, { "epoch": 0.7651572678949682, "grad_norm": 0.19337809085845947, "learning_rate": 9.05181677410131e-05, "loss": 0.0399, "step": 38290 }, { "epoch": 0.7653571000359698, "grad_norm": 0.17224670946598053, "learning_rate": 9.051211178990133e-05, "loss": 0.0444, "step": 38300 }, { "epoch": 0.7655569321769713, "grad_norm": 0.17084017395973206, "learning_rate": 9.05060541081866e-05, "loss": 0.0431, "step": 38310 }, { "epoch": 0.7657567643179729, "grad_norm": 0.12953878939151764, "learning_rate": 9.049999469612762e-05, "loss": 0.037, "step": 38320 }, { "epoch": 0.7659565964589745, "grad_norm": 0.151178240776062, "learning_rate": 9.049393355398328e-05, "loss": 0.0366, "step": 38330 }, { "epoch": 0.766156428599976, "grad_norm": 0.17654763162136078, "learning_rate": 9.048787068201248e-05, "loss": 0.0423, "step": 38340 }, { "epoch": 0.7663562607409776, "grad_norm": 0.2141922563314438, "learning_rate": 9.048180608047422e-05, "loss": 0.0362, "step": 38350 }, { "epoch": 0.7665560928819791, "grad_norm": 0.15115942060947418, "learning_rate": 9.047573974962759e-05, "loss": 0.0384, "step": 38360 }, { "epoch": 0.7667559250229807, "grad_norm": 0.1333898901939392, "learning_rate": 9.046967168973171e-05, "loss": 0.0358, "step": 38370 }, { "epoch": 0.7669557571639822, "grad_norm": 0.1304636150598526, "learning_rate": 9.046360190104579e-05, "loss": 0.0451, "step": 38380 }, { "epoch": 0.7671555893049838, "grad_norm": 0.13614852726459503, "learning_rate": 9.045753038382915e-05, "loss": 0.0412, "step": 38390 }, { "epoch": 0.7673554214459853, "grad_norm": 0.14262230694293976, "learning_rate": 9.045145713834112e-05, "loss": 0.0347, "step": 38400 }, { "epoch": 0.767555253586987, "grad_norm": 0.16776911914348602, "learning_rate": 9.044538216484115e-05, "loss": 0.032, "step": 38410 }, { "epoch": 0.7677550857279885, "grad_norm": 0.15037401020526886, "learning_rate": 9.043930546358877e-05, "loss": 0.0356, "step": 38420 }, { "epoch": 0.7679549178689901, "grad_norm": 0.15399806201457977, "learning_rate": 9.043322703484354e-05, "loss": 0.0391, "step": 38430 }, { "epoch": 0.7681547500099916, "grad_norm": 0.12551665306091309, "learning_rate": 9.042714687886512e-05, "loss": 0.0358, "step": 38440 }, { "epoch": 0.7683545821509932, "grad_norm": 0.14021138846874237, "learning_rate": 9.042106499591325e-05, "loss": 0.0405, "step": 38450 }, { "epoch": 0.7685544142919947, "grad_norm": 0.12388569116592407, "learning_rate": 9.041498138624776e-05, "loss": 0.043, "step": 38460 }, { "epoch": 0.7687542464329963, "grad_norm": 0.19070018827915192, "learning_rate": 9.04088960501285e-05, "loss": 0.0281, "step": 38470 }, { "epoch": 0.7689540785739979, "grad_norm": 0.133903369307518, "learning_rate": 9.040280898781544e-05, "loss": 0.0464, "step": 38480 }, { "epoch": 0.7691539107149994, "grad_norm": 0.12922833859920502, "learning_rate": 9.039672019956858e-05, "loss": 0.0392, "step": 38490 }, { "epoch": 0.769353742856001, "grad_norm": 0.15602637827396393, "learning_rate": 9.039062968564806e-05, "loss": 0.041, "step": 38500 }, { "epoch": 0.7695535749970025, "grad_norm": 0.11740754544734955, "learning_rate": 9.038453744631405e-05, "loss": 0.0404, "step": 38510 }, { "epoch": 0.7697534071380041, "grad_norm": 0.12633346021175385, "learning_rate": 9.037844348182678e-05, "loss": 0.0372, "step": 38520 }, { "epoch": 0.7699532392790056, "grad_norm": 0.12057612836360931, "learning_rate": 9.037234779244657e-05, "loss": 0.0369, "step": 38530 }, { "epoch": 0.7701530714200072, "grad_norm": 0.1402764618396759, "learning_rate": 9.036625037843384e-05, "loss": 0.0347, "step": 38540 }, { "epoch": 0.7703529035610087, "grad_norm": 0.14664769172668457, "learning_rate": 9.036015124004906e-05, "loss": 0.0435, "step": 38550 }, { "epoch": 0.7705527357020103, "grad_norm": 0.1487794816493988, "learning_rate": 9.035405037755275e-05, "loss": 0.0412, "step": 38560 }, { "epoch": 0.7707525678430118, "grad_norm": 0.13429129123687744, "learning_rate": 9.034794779120554e-05, "loss": 0.0464, "step": 38570 }, { "epoch": 0.7709523999840134, "grad_norm": 0.13337507843971252, "learning_rate": 9.034184348126813e-05, "loss": 0.0389, "step": 38580 }, { "epoch": 0.7711522321250149, "grad_norm": 0.12136750668287277, "learning_rate": 9.033573744800127e-05, "loss": 0.0396, "step": 38590 }, { "epoch": 0.7713520642660165, "grad_norm": 0.14150108397006989, "learning_rate": 9.03296296916658e-05, "loss": 0.0357, "step": 38600 }, { "epoch": 0.7715518964070182, "grad_norm": 0.15661026537418365, "learning_rate": 9.032352021252264e-05, "loss": 0.0329, "step": 38610 }, { "epoch": 0.7717517285480197, "grad_norm": 0.14790937304496765, "learning_rate": 9.03174090108328e-05, "loss": 0.0382, "step": 38620 }, { "epoch": 0.7719515606890213, "grad_norm": 0.11379306018352509, "learning_rate": 9.031129608685727e-05, "loss": 0.0342, "step": 38630 }, { "epoch": 0.7721513928300228, "grad_norm": 0.1366858333349228, "learning_rate": 9.030518144085725e-05, "loss": 0.0334, "step": 38640 }, { "epoch": 0.7723512249710244, "grad_norm": 0.1330738663673401, "learning_rate": 9.029906507309391e-05, "loss": 0.0357, "step": 38650 }, { "epoch": 0.7725510571120259, "grad_norm": 0.12232456356287003, "learning_rate": 9.029294698382855e-05, "loss": 0.039, "step": 38660 }, { "epoch": 0.7727508892530275, "grad_norm": 0.11894690245389938, "learning_rate": 9.028682717332254e-05, "loss": 0.042, "step": 38670 }, { "epoch": 0.772950721394029, "grad_norm": 0.16276361048221588, "learning_rate": 9.028070564183724e-05, "loss": 0.0347, "step": 38680 }, { "epoch": 0.7731505535350306, "grad_norm": 0.15407602488994598, "learning_rate": 9.027458238963421e-05, "loss": 0.0338, "step": 38690 }, { "epoch": 0.7733503856760321, "grad_norm": 0.2058417797088623, "learning_rate": 9.026845741697501e-05, "loss": 0.0444, "step": 38700 }, { "epoch": 0.7735502178170337, "grad_norm": 0.17398589849472046, "learning_rate": 9.026233072412128e-05, "loss": 0.0347, "step": 38710 }, { "epoch": 0.7737500499580352, "grad_norm": 0.18768341839313507, "learning_rate": 9.025620231133474e-05, "loss": 0.0334, "step": 38720 }, { "epoch": 0.7739498820990368, "grad_norm": 0.14090070128440857, "learning_rate": 9.02500721788772e-05, "loss": 0.0345, "step": 38730 }, { "epoch": 0.7741497142400384, "grad_norm": 0.18453694880008698, "learning_rate": 9.024394032701051e-05, "loss": 0.039, "step": 38740 }, { "epoch": 0.7743495463810399, "grad_norm": 0.1373310387134552, "learning_rate": 9.023780675599664e-05, "loss": 0.0408, "step": 38750 }, { "epoch": 0.7745493785220415, "grad_norm": 0.12930692732334137, "learning_rate": 9.023167146609756e-05, "loss": 0.0401, "step": 38760 }, { "epoch": 0.774749210663043, "grad_norm": 0.12451305240392685, "learning_rate": 9.02255344575754e-05, "loss": 0.033, "step": 38770 }, { "epoch": 0.7749490428040446, "grad_norm": 0.15587995946407318, "learning_rate": 9.021939573069229e-05, "loss": 0.0465, "step": 38780 }, { "epoch": 0.7751488749450461, "grad_norm": 0.15120546519756317, "learning_rate": 9.02132552857105e-05, "loss": 0.0373, "step": 38790 }, { "epoch": 0.7753487070860478, "grad_norm": 0.18891283869743347, "learning_rate": 9.020711312289229e-05, "loss": 0.0358, "step": 38800 }, { "epoch": 0.7755485392270492, "grad_norm": 0.19163738191127777, "learning_rate": 9.02009692425001e-05, "loss": 0.0377, "step": 38810 }, { "epoch": 0.7757483713680509, "grad_norm": 0.16640131175518036, "learning_rate": 9.019482364479634e-05, "loss": 0.0463, "step": 38820 }, { "epoch": 0.7759482035090524, "grad_norm": 0.16403770446777344, "learning_rate": 9.018867633004355e-05, "loss": 0.0378, "step": 38830 }, { "epoch": 0.776148035650054, "grad_norm": 0.14587394893169403, "learning_rate": 9.018252729850434e-05, "loss": 0.0341, "step": 38840 }, { "epoch": 0.7763478677910555, "grad_norm": 0.16550014913082123, "learning_rate": 9.01763765504414e-05, "loss": 0.0343, "step": 38850 }, { "epoch": 0.7765476999320571, "grad_norm": 0.16843357682228088, "learning_rate": 9.017022408611745e-05, "loss": 0.04, "step": 38860 }, { "epoch": 0.7767475320730586, "grad_norm": 0.13321805000305176, "learning_rate": 9.016406990579529e-05, "loss": 0.0382, "step": 38870 }, { "epoch": 0.7769473642140602, "grad_norm": 0.1703367531299591, "learning_rate": 9.015791400973787e-05, "loss": 0.0434, "step": 38880 }, { "epoch": 0.7771471963550618, "grad_norm": 0.19754619896411896, "learning_rate": 9.015175639820814e-05, "loss": 0.0421, "step": 38890 }, { "epoch": 0.7773470284960633, "grad_norm": 0.12908132374286652, "learning_rate": 9.014559707146914e-05, "loss": 0.0316, "step": 38900 }, { "epoch": 0.7775468606370649, "grad_norm": 0.11615409702062607, "learning_rate": 9.013943602978399e-05, "loss": 0.036, "step": 38910 }, { "epoch": 0.7777466927780664, "grad_norm": 0.13190770149230957, "learning_rate": 9.013327327341586e-05, "loss": 0.0333, "step": 38920 }, { "epoch": 0.777946524919068, "grad_norm": 0.18129347264766693, "learning_rate": 9.012710880262803e-05, "loss": 0.0436, "step": 38930 }, { "epoch": 0.7781463570600695, "grad_norm": 0.13682112097740173, "learning_rate": 9.01209426176838e-05, "loss": 0.0391, "step": 38940 }, { "epoch": 0.7783461892010711, "grad_norm": 0.10866957902908325, "learning_rate": 9.011477471884664e-05, "loss": 0.0497, "step": 38950 }, { "epoch": 0.7785460213420726, "grad_norm": 0.13575826585292816, "learning_rate": 9.010860510637997e-05, "loss": 0.0387, "step": 38960 }, { "epoch": 0.7787458534830742, "grad_norm": 0.20636554062366486, "learning_rate": 9.01024337805474e-05, "loss": 0.0615, "step": 38970 }, { "epoch": 0.7789456856240757, "grad_norm": 0.16300475597381592, "learning_rate": 9.009626074161252e-05, "loss": 0.032, "step": 38980 }, { "epoch": 0.7791455177650773, "grad_norm": 0.14306795597076416, "learning_rate": 9.009008598983903e-05, "loss": 0.0344, "step": 38990 }, { "epoch": 0.7793453499060788, "grad_norm": 0.1626938283443451, "learning_rate": 9.008390952549071e-05, "loss": 0.037, "step": 39000 }, { "epoch": 0.7795451820470805, "grad_norm": 0.1460154801607132, "learning_rate": 9.007773134883144e-05, "loss": 0.0343, "step": 39010 }, { "epoch": 0.7797450141880821, "grad_norm": 0.10404793918132782, "learning_rate": 9.00715514601251e-05, "loss": 0.0376, "step": 39020 }, { "epoch": 0.7799448463290836, "grad_norm": 0.14253416657447815, "learning_rate": 9.006536985963567e-05, "loss": 0.0402, "step": 39030 }, { "epoch": 0.7801446784700852, "grad_norm": 0.1568298190832138, "learning_rate": 9.005918654762727e-05, "loss": 0.031, "step": 39040 }, { "epoch": 0.7803445106110867, "grad_norm": 0.16236785054206848, "learning_rate": 9.0053001524364e-05, "loss": 0.0436, "step": 39050 }, { "epoch": 0.7805443427520883, "grad_norm": 0.1381082385778427, "learning_rate": 9.004681479011008e-05, "loss": 0.0376, "step": 39060 }, { "epoch": 0.7807441748930898, "grad_norm": 0.15190915763378143, "learning_rate": 9.004062634512981e-05, "loss": 0.0362, "step": 39070 }, { "epoch": 0.7809440070340914, "grad_norm": 0.16825716197490692, "learning_rate": 9.003443618968754e-05, "loss": 0.0396, "step": 39080 }, { "epoch": 0.7811438391750929, "grad_norm": 0.1453905552625656, "learning_rate": 9.00282443240477e-05, "loss": 0.0329, "step": 39090 }, { "epoch": 0.7813436713160945, "grad_norm": 0.10574153065681458, "learning_rate": 9.00220507484748e-05, "loss": 0.0484, "step": 39100 }, { "epoch": 0.781543503457096, "grad_norm": 0.15078726410865784, "learning_rate": 9.001585546323341e-05, "loss": 0.0407, "step": 39110 }, { "epoch": 0.7817433355980976, "grad_norm": 0.1551014482975006, "learning_rate": 9.000965846858819e-05, "loss": 0.0382, "step": 39120 }, { "epoch": 0.7819431677390991, "grad_norm": 0.13952165842056274, "learning_rate": 9.000345976480385e-05, "loss": 0.0484, "step": 39130 }, { "epoch": 0.7821429998801007, "grad_norm": 0.13071322441101074, "learning_rate": 8.99972593521452e-05, "loss": 0.0404, "step": 39140 }, { "epoch": 0.7823428320211022, "grad_norm": 0.13956420123577118, "learning_rate": 8.999105723087713e-05, "loss": 0.0363, "step": 39150 }, { "epoch": 0.7825426641621038, "grad_norm": 0.15806733071804047, "learning_rate": 8.998485340126453e-05, "loss": 0.0406, "step": 39160 }, { "epoch": 0.7827424963031054, "grad_norm": 0.1290970891714096, "learning_rate": 8.997864786357247e-05, "loss": 0.0308, "step": 39170 }, { "epoch": 0.7829423284441069, "grad_norm": 0.11408352851867676, "learning_rate": 8.9972440618066e-05, "loss": 0.0371, "step": 39180 }, { "epoch": 0.7831421605851085, "grad_norm": 0.10727600008249283, "learning_rate": 8.996623166501031e-05, "loss": 0.0309, "step": 39190 }, { "epoch": 0.78334199272611, "grad_norm": 0.11435763537883759, "learning_rate": 8.996002100467063e-05, "loss": 0.0419, "step": 39200 }, { "epoch": 0.7835418248671117, "grad_norm": 0.1130562424659729, "learning_rate": 8.995380863731225e-05, "loss": 0.0346, "step": 39210 }, { "epoch": 0.7837416570081132, "grad_norm": 0.12857821583747864, "learning_rate": 8.994759456320055e-05, "loss": 0.0327, "step": 39220 }, { "epoch": 0.7839414891491148, "grad_norm": 0.11880967020988464, "learning_rate": 8.994137878260101e-05, "loss": 0.0355, "step": 39230 }, { "epoch": 0.7841413212901163, "grad_norm": 0.15423588454723358, "learning_rate": 8.993516129577914e-05, "loss": 0.0378, "step": 39240 }, { "epoch": 0.7843411534311179, "grad_norm": 0.11563505977392197, "learning_rate": 8.992894210300056e-05, "loss": 0.0412, "step": 39250 }, { "epoch": 0.7845409855721194, "grad_norm": 0.12702840566635132, "learning_rate": 8.992272120453091e-05, "loss": 0.0329, "step": 39260 }, { "epoch": 0.784740817713121, "grad_norm": 0.13275860249996185, "learning_rate": 8.991649860063594e-05, "loss": 0.036, "step": 39270 }, { "epoch": 0.7849406498541225, "grad_norm": 0.12815320491790771, "learning_rate": 8.99102742915815e-05, "loss": 0.0307, "step": 39280 }, { "epoch": 0.7851404819951241, "grad_norm": 0.1517430990934372, "learning_rate": 8.990404827763344e-05, "loss": 0.0293, "step": 39290 }, { "epoch": 0.7853403141361257, "grad_norm": 0.13319993019104004, "learning_rate": 8.989782055905775e-05, "loss": 0.0362, "step": 39300 }, { "epoch": 0.7855401462771272, "grad_norm": 0.14218993484973907, "learning_rate": 8.989159113612047e-05, "loss": 0.0412, "step": 39310 }, { "epoch": 0.7857399784181288, "grad_norm": 0.162324920296669, "learning_rate": 8.98853600090877e-05, "loss": 0.0457, "step": 39320 }, { "epoch": 0.7859398105591303, "grad_norm": 0.1302458494901657, "learning_rate": 8.987912717822562e-05, "loss": 0.0446, "step": 39330 }, { "epoch": 0.7861396427001319, "grad_norm": 0.17507995665073395, "learning_rate": 8.987289264380048e-05, "loss": 0.0419, "step": 39340 }, { "epoch": 0.7863394748411334, "grad_norm": 0.15373101830482483, "learning_rate": 8.98666564060786e-05, "loss": 0.0301, "step": 39350 }, { "epoch": 0.786539306982135, "grad_norm": 0.14084842801094055, "learning_rate": 8.986041846532642e-05, "loss": 0.0373, "step": 39360 }, { "epoch": 0.7867391391231365, "grad_norm": 0.17465811967849731, "learning_rate": 8.98541788218104e-05, "loss": 0.0343, "step": 39370 }, { "epoch": 0.7869389712641381, "grad_norm": 0.14208555221557617, "learning_rate": 8.984793747579705e-05, "loss": 0.0358, "step": 39380 }, { "epoch": 0.7871388034051396, "grad_norm": 0.11430330574512482, "learning_rate": 8.984169442755301e-05, "loss": 0.0395, "step": 39390 }, { "epoch": 0.7873386355461413, "grad_norm": 0.15893295407295227, "learning_rate": 8.9835449677345e-05, "loss": 0.045, "step": 39400 }, { "epoch": 0.7875384676871428, "grad_norm": 0.12314515560865402, "learning_rate": 8.982920322543976e-05, "loss": 0.0319, "step": 39410 }, { "epoch": 0.7877382998281444, "grad_norm": 0.11311404407024384, "learning_rate": 8.98229550721041e-05, "loss": 0.0401, "step": 39420 }, { "epoch": 0.787938131969146, "grad_norm": 0.14936424791812897, "learning_rate": 8.981670521760497e-05, "loss": 0.0433, "step": 39430 }, { "epoch": 0.7881379641101475, "grad_norm": 0.1448599100112915, "learning_rate": 8.981045366220934e-05, "loss": 0.0368, "step": 39440 }, { "epoch": 0.7883377962511491, "grad_norm": 0.12688012421131134, "learning_rate": 8.980420040618427e-05, "loss": 0.0468, "step": 39450 }, { "epoch": 0.7885376283921506, "grad_norm": 0.15547090768814087, "learning_rate": 8.979794544979687e-05, "loss": 0.0304, "step": 39460 }, { "epoch": 0.7887374605331522, "grad_norm": 0.11412856727838516, "learning_rate": 8.979168879331436e-05, "loss": 0.0391, "step": 39470 }, { "epoch": 0.7889372926741537, "grad_norm": 0.15981703996658325, "learning_rate": 8.978543043700399e-05, "loss": 0.0414, "step": 39480 }, { "epoch": 0.7891371248151553, "grad_norm": 0.1698087602853775, "learning_rate": 8.977917038113312e-05, "loss": 0.0476, "step": 39490 }, { "epoch": 0.7893369569561568, "grad_norm": 0.17456327378749847, "learning_rate": 8.977290862596919e-05, "loss": 0.0316, "step": 39500 }, { "epoch": 0.7895367890971584, "grad_norm": 0.12529008090496063, "learning_rate": 8.976664517177966e-05, "loss": 0.0447, "step": 39510 }, { "epoch": 0.7897366212381599, "grad_norm": 0.14476102590560913, "learning_rate": 8.97603800188321e-05, "loss": 0.033, "step": 39520 }, { "epoch": 0.7899364533791615, "grad_norm": 0.16204334795475006, "learning_rate": 8.975411316739415e-05, "loss": 0.0432, "step": 39530 }, { "epoch": 0.790136285520163, "grad_norm": 0.1398642659187317, "learning_rate": 8.974784461773351e-05, "loss": 0.0409, "step": 39540 }, { "epoch": 0.7903361176611646, "grad_norm": 0.1851595789194107, "learning_rate": 8.974157437011796e-05, "loss": 0.0422, "step": 39550 }, { "epoch": 0.7905359498021661, "grad_norm": 0.18786557018756866, "learning_rate": 8.973530242481538e-05, "loss": 0.033, "step": 39560 }, { "epoch": 0.7907357819431677, "grad_norm": 0.18444983661174774, "learning_rate": 8.972902878209366e-05, "loss": 0.0392, "step": 39570 }, { "epoch": 0.7909356140841693, "grad_norm": 0.18764150142669678, "learning_rate": 8.972275344222084e-05, "loss": 0.0408, "step": 39580 }, { "epoch": 0.7911354462251708, "grad_norm": 0.13908268511295319, "learning_rate": 8.971647640546493e-05, "loss": 0.0373, "step": 39590 }, { "epoch": 0.7913352783661725, "grad_norm": 0.14684540033340454, "learning_rate": 8.971019767209414e-05, "loss": 0.0402, "step": 39600 }, { "epoch": 0.791535110507174, "grad_norm": 0.1316988468170166, "learning_rate": 8.970391724237663e-05, "loss": 0.0453, "step": 39610 }, { "epoch": 0.7917349426481756, "grad_norm": 0.125022754073143, "learning_rate": 8.969763511658074e-05, "loss": 0.0357, "step": 39620 }, { "epoch": 0.7919347747891771, "grad_norm": 0.15404266119003296, "learning_rate": 8.969135129497479e-05, "loss": 0.0299, "step": 39630 }, { "epoch": 0.7921346069301787, "grad_norm": 0.2097882628440857, "learning_rate": 8.968506577782723e-05, "loss": 0.0436, "step": 39640 }, { "epoch": 0.7923344390711802, "grad_norm": 0.1751798540353775, "learning_rate": 8.967877856540657e-05, "loss": 0.0461, "step": 39650 }, { "epoch": 0.7925342712121818, "grad_norm": 0.14847798645496368, "learning_rate": 8.967248965798136e-05, "loss": 0.034, "step": 39660 }, { "epoch": 0.7927341033531833, "grad_norm": 0.12996487319469452, "learning_rate": 8.96661990558203e-05, "loss": 0.0425, "step": 39670 }, { "epoch": 0.7929339354941849, "grad_norm": 0.16432301700115204, "learning_rate": 8.965990675919207e-05, "loss": 0.041, "step": 39680 }, { "epoch": 0.7931337676351864, "grad_norm": 0.16960689425468445, "learning_rate": 8.965361276836549e-05, "loss": 0.0418, "step": 39690 }, { "epoch": 0.793333599776188, "grad_norm": 0.18280227482318878, "learning_rate": 8.96473170836094e-05, "loss": 0.0489, "step": 39700 }, { "epoch": 0.7935334319171896, "grad_norm": 0.14284014701843262, "learning_rate": 8.964101970519276e-05, "loss": 0.0447, "step": 39710 }, { "epoch": 0.7937332640581911, "grad_norm": 0.11825637519359589, "learning_rate": 8.96347206333846e-05, "loss": 0.0442, "step": 39720 }, { "epoch": 0.7939330961991927, "grad_norm": 0.12983161211013794, "learning_rate": 8.962841986845399e-05, "loss": 0.0384, "step": 39730 }, { "epoch": 0.7941329283401942, "grad_norm": 0.16382841765880585, "learning_rate": 8.962211741067006e-05, "loss": 0.0431, "step": 39740 }, { "epoch": 0.7943327604811958, "grad_norm": 0.1336696743965149, "learning_rate": 8.961581326030207e-05, "loss": 0.0342, "step": 39750 }, { "epoch": 0.7945325926221973, "grad_norm": 0.15752068161964417, "learning_rate": 8.960950741761932e-05, "loss": 0.0343, "step": 39760 }, { "epoch": 0.7947324247631989, "grad_norm": 0.1352953165769577, "learning_rate": 8.960319988289119e-05, "loss": 0.0394, "step": 39770 }, { "epoch": 0.7949322569042004, "grad_norm": 0.12422606348991394, "learning_rate": 8.959689065638709e-05, "loss": 0.0393, "step": 39780 }, { "epoch": 0.795132089045202, "grad_norm": 0.13734503090381622, "learning_rate": 8.959057973837657e-05, "loss": 0.0372, "step": 39790 }, { "epoch": 0.7953319211862035, "grad_norm": 0.12660396099090576, "learning_rate": 8.958426712912922e-05, "loss": 0.0385, "step": 39800 }, { "epoch": 0.7955317533272052, "grad_norm": 0.13086719810962677, "learning_rate": 8.957795282891469e-05, "loss": 0.0297, "step": 39810 }, { "epoch": 0.7957315854682067, "grad_norm": 0.13881012797355652, "learning_rate": 8.957163683800274e-05, "loss": 0.0392, "step": 39820 }, { "epoch": 0.7959314176092083, "grad_norm": 0.14520341157913208, "learning_rate": 8.956531915666313e-05, "loss": 0.0374, "step": 39830 }, { "epoch": 0.7961312497502098, "grad_norm": 0.13737048208713531, "learning_rate": 8.955899978516579e-05, "loss": 0.0355, "step": 39840 }, { "epoch": 0.7963310818912114, "grad_norm": 0.12454380095005035, "learning_rate": 8.955267872378064e-05, "loss": 0.0327, "step": 39850 }, { "epoch": 0.796530914032213, "grad_norm": 0.11585891991853714, "learning_rate": 8.954635597277772e-05, "loss": 0.036, "step": 39860 }, { "epoch": 0.7967307461732145, "grad_norm": 0.11149203032255173, "learning_rate": 8.954003153242713e-05, "loss": 0.0319, "step": 39870 }, { "epoch": 0.7969305783142161, "grad_norm": 0.12755528092384338, "learning_rate": 8.953370540299902e-05, "loss": 0.0398, "step": 39880 }, { "epoch": 0.7971304104552176, "grad_norm": 0.16190622746944427, "learning_rate": 8.952737758476365e-05, "loss": 0.0347, "step": 39890 }, { "epoch": 0.7973302425962192, "grad_norm": 0.1264229267835617, "learning_rate": 8.952104807799131e-05, "loss": 0.0429, "step": 39900 }, { "epoch": 0.7975300747372207, "grad_norm": 0.13874773681163788, "learning_rate": 8.951471688295241e-05, "loss": 0.0321, "step": 39910 }, { "epoch": 0.7977299068782223, "grad_norm": 0.1435568779706955, "learning_rate": 8.95083839999174e-05, "loss": 0.0411, "step": 39920 }, { "epoch": 0.7979297390192238, "grad_norm": 0.14032554626464844, "learning_rate": 8.950204942915678e-05, "loss": 0.039, "step": 39930 }, { "epoch": 0.7981295711602254, "grad_norm": 0.3597816228866577, "learning_rate": 8.94957131709412e-05, "loss": 0.0415, "step": 39940 }, { "epoch": 0.7983294033012269, "grad_norm": 0.13903924822807312, "learning_rate": 8.948937522554131e-05, "loss": 0.0299, "step": 39950 }, { "epoch": 0.7985292354422285, "grad_norm": 0.12427831441164017, "learning_rate": 8.948303559322784e-05, "loss": 0.0379, "step": 39960 }, { "epoch": 0.79872906758323, "grad_norm": 0.159464031457901, "learning_rate": 8.947669427427163e-05, "loss": 0.0401, "step": 39970 }, { "epoch": 0.7989288997242316, "grad_norm": 0.16474243998527527, "learning_rate": 8.947035126894356e-05, "loss": 0.0428, "step": 39980 }, { "epoch": 0.7991287318652333, "grad_norm": 0.12607745826244354, "learning_rate": 8.94640065775146e-05, "loss": 0.0336, "step": 39990 }, { "epoch": 0.7993285640062348, "grad_norm": 0.1473509669303894, "learning_rate": 8.945766020025577e-05, "loss": 0.0383, "step": 40000 }, { "epoch": 0.7995283961472364, "grad_norm": 0.12119954824447632, "learning_rate": 8.945131213743818e-05, "loss": 0.0366, "step": 40010 }, { "epoch": 0.7997282282882379, "grad_norm": 0.1481960117816925, "learning_rate": 8.944496238933303e-05, "loss": 0.0364, "step": 40020 }, { "epoch": 0.7999280604292395, "grad_norm": 0.10495302081108093, "learning_rate": 8.943861095621152e-05, "loss": 0.0366, "step": 40030 }, { "epoch": 0.800127892570241, "grad_norm": 0.1482575535774231, "learning_rate": 8.943225783834501e-05, "loss": 0.0336, "step": 40040 }, { "epoch": 0.8003277247112426, "grad_norm": 0.16157828271389008, "learning_rate": 8.942590303600489e-05, "loss": 0.0396, "step": 40050 }, { "epoch": 0.8005275568522441, "grad_norm": 0.11624351888895035, "learning_rate": 8.941954654946264e-05, "loss": 0.037, "step": 40060 }, { "epoch": 0.8007273889932457, "grad_norm": 0.10222730785608292, "learning_rate": 8.941318837898973e-05, "loss": 0.0324, "step": 40070 }, { "epoch": 0.8009272211342472, "grad_norm": 0.16645093262195587, "learning_rate": 8.940682852485784e-05, "loss": 0.042, "step": 40080 }, { "epoch": 0.8011270532752488, "grad_norm": 0.10407771915197372, "learning_rate": 8.940046698733864e-05, "loss": 0.0329, "step": 40090 }, { "epoch": 0.8013268854162503, "grad_norm": 0.11148910224437714, "learning_rate": 8.939410376670387e-05, "loss": 0.0386, "step": 40100 }, { "epoch": 0.8015267175572519, "grad_norm": 0.11906757950782776, "learning_rate": 8.938773886322534e-05, "loss": 0.0426, "step": 40110 }, { "epoch": 0.8017265496982535, "grad_norm": 0.12103618681430817, "learning_rate": 8.938137227717498e-05, "loss": 0.0348, "step": 40120 }, { "epoch": 0.801926381839255, "grad_norm": 0.13862308859825134, "learning_rate": 8.937500400882473e-05, "loss": 0.0361, "step": 40130 }, { "epoch": 0.8021262139802566, "grad_norm": 0.14143399894237518, "learning_rate": 8.936863405844665e-05, "loss": 0.0377, "step": 40140 }, { "epoch": 0.8023260461212581, "grad_norm": 0.17462888360023499, "learning_rate": 8.936226242631282e-05, "loss": 0.0358, "step": 40150 }, { "epoch": 0.8025258782622597, "grad_norm": 0.11373192071914673, "learning_rate": 8.935588911269548e-05, "loss": 0.0325, "step": 40160 }, { "epoch": 0.8027257104032612, "grad_norm": 0.14630140364170074, "learning_rate": 8.934951411786685e-05, "loss": 0.0325, "step": 40170 }, { "epoch": 0.8029255425442628, "grad_norm": 0.12107110768556595, "learning_rate": 8.934313744209926e-05, "loss": 0.0438, "step": 40180 }, { "epoch": 0.8031253746852643, "grad_norm": 0.1455579400062561, "learning_rate": 8.933675908566512e-05, "loss": 0.0409, "step": 40190 }, { "epoch": 0.803325206826266, "grad_norm": 0.15347526967525482, "learning_rate": 8.93303790488369e-05, "loss": 0.0336, "step": 40200 }, { "epoch": 0.8035250389672675, "grad_norm": 0.17003268003463745, "learning_rate": 8.932399733188712e-05, "loss": 0.0416, "step": 40210 }, { "epoch": 0.8037248711082691, "grad_norm": 0.13223910331726074, "learning_rate": 8.931761393508843e-05, "loss": 0.028, "step": 40220 }, { "epoch": 0.8039247032492706, "grad_norm": 0.12478466331958771, "learning_rate": 8.93112288587135e-05, "loss": 0.0316, "step": 40230 }, { "epoch": 0.8041245353902722, "grad_norm": 0.13799338042736053, "learning_rate": 8.930484210303509e-05, "loss": 0.0386, "step": 40240 }, { "epoch": 0.8043243675312737, "grad_norm": 0.13253623247146606, "learning_rate": 8.929845366832603e-05, "loss": 0.0306, "step": 40250 }, { "epoch": 0.8045241996722753, "grad_norm": 0.17109133303165436, "learning_rate": 8.929206355485922e-05, "loss": 0.0372, "step": 40260 }, { "epoch": 0.8047240318132769, "grad_norm": 0.1405564844608307, "learning_rate": 8.928567176290763e-05, "loss": 0.043, "step": 40270 }, { "epoch": 0.8049238639542784, "grad_norm": 0.12874184548854828, "learning_rate": 8.927927829274433e-05, "loss": 0.0418, "step": 40280 }, { "epoch": 0.80512369609528, "grad_norm": 0.15681257843971252, "learning_rate": 8.92728831446424e-05, "loss": 0.0382, "step": 40290 }, { "epoch": 0.8053235282362815, "grad_norm": 0.1445283591747284, "learning_rate": 8.926648631887506e-05, "loss": 0.0429, "step": 40300 }, { "epoch": 0.8055233603772831, "grad_norm": 0.15228602290153503, "learning_rate": 8.926008781571555e-05, "loss": 0.0431, "step": 40310 }, { "epoch": 0.8057231925182846, "grad_norm": 0.16332994401454926, "learning_rate": 8.925368763543723e-05, "loss": 0.0414, "step": 40320 }, { "epoch": 0.8059230246592862, "grad_norm": 0.16958807408809662, "learning_rate": 8.924728577831347e-05, "loss": 0.0321, "step": 40330 }, { "epoch": 0.8061228568002877, "grad_norm": 0.15084408223628998, "learning_rate": 8.924088224461777e-05, "loss": 0.0381, "step": 40340 }, { "epoch": 0.8063226889412893, "grad_norm": 0.15834219753742218, "learning_rate": 8.923447703462367e-05, "loss": 0.0403, "step": 40350 }, { "epoch": 0.8065225210822908, "grad_norm": 0.14792363345623016, "learning_rate": 8.922807014860479e-05, "loss": 0.0386, "step": 40360 }, { "epoch": 0.8067223532232924, "grad_norm": 0.1387052834033966, "learning_rate": 8.92216615868348e-05, "loss": 0.0367, "step": 40370 }, { "epoch": 0.8069221853642939, "grad_norm": 0.17198878526687622, "learning_rate": 8.921525134958748e-05, "loss": 0.0396, "step": 40380 }, { "epoch": 0.8071220175052956, "grad_norm": 0.15214332938194275, "learning_rate": 8.920883943713667e-05, "loss": 0.0285, "step": 40390 }, { "epoch": 0.8073218496462972, "grad_norm": 0.14761607348918915, "learning_rate": 8.920242584975627e-05, "loss": 0.0552, "step": 40400 }, { "epoch": 0.8075216817872987, "grad_norm": 0.13165900111198425, "learning_rate": 8.919601058772027e-05, "loss": 0.0426, "step": 40410 }, { "epoch": 0.8077215139283003, "grad_norm": 0.13141953945159912, "learning_rate": 8.918959365130268e-05, "loss": 0.0359, "step": 40420 }, { "epoch": 0.8079213460693018, "grad_norm": 0.1321757733821869, "learning_rate": 8.918317504077764e-05, "loss": 0.0358, "step": 40430 }, { "epoch": 0.8081211782103034, "grad_norm": 0.14046752452850342, "learning_rate": 8.917675475641937e-05, "loss": 0.0468, "step": 40440 }, { "epoch": 0.8083210103513049, "grad_norm": 0.1531643271446228, "learning_rate": 8.917033279850208e-05, "loss": 0.0328, "step": 40450 }, { "epoch": 0.8085208424923065, "grad_norm": 0.23019316792488098, "learning_rate": 8.916390916730015e-05, "loss": 0.032, "step": 40460 }, { "epoch": 0.808720674633308, "grad_norm": 0.16142143309116364, "learning_rate": 8.915748386308796e-05, "loss": 0.0389, "step": 40470 }, { "epoch": 0.8089205067743096, "grad_norm": 0.13439315557479858, "learning_rate": 8.915105688613999e-05, "loss": 0.0451, "step": 40480 }, { "epoch": 0.8091203389153111, "grad_norm": 0.18770632147789001, "learning_rate": 8.91446282367308e-05, "loss": 0.0316, "step": 40490 }, { "epoch": 0.8093201710563127, "grad_norm": 0.13111332058906555, "learning_rate": 8.9138197915135e-05, "loss": 0.0291, "step": 40500 }, { "epoch": 0.8095200031973142, "grad_norm": 0.11711510270833969, "learning_rate": 8.913176592162728e-05, "loss": 0.0411, "step": 40510 }, { "epoch": 0.8097198353383158, "grad_norm": 0.13778600096702576, "learning_rate": 8.91253322564824e-05, "loss": 0.034, "step": 40520 }, { "epoch": 0.8099196674793174, "grad_norm": 0.17295384407043457, "learning_rate": 8.911889691997523e-05, "loss": 0.0457, "step": 40530 }, { "epoch": 0.8101194996203189, "grad_norm": 0.14383523166179657, "learning_rate": 8.911245991238062e-05, "loss": 0.0374, "step": 40540 }, { "epoch": 0.8103193317613205, "grad_norm": 0.14467404782772064, "learning_rate": 8.910602123397357e-05, "loss": 0.0327, "step": 40550 }, { "epoch": 0.810519163902322, "grad_norm": 0.12877127528190613, "learning_rate": 8.909958088502916e-05, "loss": 0.0364, "step": 40560 }, { "epoch": 0.8107189960433236, "grad_norm": 0.1565353274345398, "learning_rate": 8.909313886582247e-05, "loss": 0.0298, "step": 40570 }, { "epoch": 0.8109188281843251, "grad_norm": 0.1312144696712494, "learning_rate": 8.908669517662871e-05, "loss": 0.0354, "step": 40580 }, { "epoch": 0.8111186603253268, "grad_norm": 0.16766439378261566, "learning_rate": 8.908024981772312e-05, "loss": 0.0414, "step": 40590 }, { "epoch": 0.8113184924663283, "grad_norm": 0.10972180217504501, "learning_rate": 8.907380278938107e-05, "loss": 0.0292, "step": 40600 }, { "epoch": 0.8115183246073299, "grad_norm": 0.15509481728076935, "learning_rate": 8.906735409187792e-05, "loss": 0.0356, "step": 40610 }, { "epoch": 0.8117181567483314, "grad_norm": 0.131755068898201, "learning_rate": 8.90609037254892e-05, "loss": 0.0363, "step": 40620 }, { "epoch": 0.811917988889333, "grad_norm": 0.10359352082014084, "learning_rate": 8.905445169049041e-05, "loss": 0.0365, "step": 40630 }, { "epoch": 0.8121178210303345, "grad_norm": 0.12450318783521652, "learning_rate": 8.90479979871572e-05, "loss": 0.0329, "step": 40640 }, { "epoch": 0.8123176531713361, "grad_norm": 0.11990872770547867, "learning_rate": 8.904154261576524e-05, "loss": 0.0422, "step": 40650 }, { "epoch": 0.8125174853123376, "grad_norm": 0.12152686715126038, "learning_rate": 8.903508557659032e-05, "loss": 0.037, "step": 40660 }, { "epoch": 0.8127173174533392, "grad_norm": 0.150826558470726, "learning_rate": 8.902862686990824e-05, "loss": 0.0423, "step": 40670 }, { "epoch": 0.8129171495943408, "grad_norm": 0.1453421264886856, "learning_rate": 8.902216649599492e-05, "loss": 0.0534, "step": 40680 }, { "epoch": 0.8131169817353423, "grad_norm": 0.1270623803138733, "learning_rate": 8.901570445512633e-05, "loss": 0.0361, "step": 40690 }, { "epoch": 0.8133168138763439, "grad_norm": 0.19250428676605225, "learning_rate": 8.900924074757853e-05, "loss": 0.0372, "step": 40700 }, { "epoch": 0.8135166460173454, "grad_norm": 0.15884657204151154, "learning_rate": 8.900277537362761e-05, "loss": 0.0423, "step": 40710 }, { "epoch": 0.813716478158347, "grad_norm": 0.15336152911186218, "learning_rate": 8.89963083335498e-05, "loss": 0.0377, "step": 40720 }, { "epoch": 0.8139163102993485, "grad_norm": 0.1324758231639862, "learning_rate": 8.898983962762132e-05, "loss": 0.0318, "step": 40730 }, { "epoch": 0.8141161424403501, "grad_norm": 0.12424296140670776, "learning_rate": 8.898336925611852e-05, "loss": 0.0386, "step": 40740 }, { "epoch": 0.8143159745813516, "grad_norm": 0.1337902843952179, "learning_rate": 8.897689721931779e-05, "loss": 0.0336, "step": 40750 }, { "epoch": 0.8145158067223532, "grad_norm": 0.1270369291305542, "learning_rate": 8.897042351749562e-05, "loss": 0.0336, "step": 40760 }, { "epoch": 0.8147156388633547, "grad_norm": 0.1632169783115387, "learning_rate": 8.896394815092856e-05, "loss": 0.0405, "step": 40770 }, { "epoch": 0.8149154710043564, "grad_norm": 0.14302556216716766, "learning_rate": 8.895747111989321e-05, "loss": 0.0366, "step": 40780 }, { "epoch": 0.8151153031453579, "grad_norm": 0.15764562785625458, "learning_rate": 8.895099242466626e-05, "loss": 0.0323, "step": 40790 }, { "epoch": 0.8153151352863595, "grad_norm": 0.16222050786018372, "learning_rate": 8.894451206552447e-05, "loss": 0.0346, "step": 40800 }, { "epoch": 0.8155149674273611, "grad_norm": 0.12648336589336395, "learning_rate": 8.893803004274465e-05, "loss": 0.0338, "step": 40810 }, { "epoch": 0.8157147995683626, "grad_norm": 0.14699362218379974, "learning_rate": 8.893154635660375e-05, "loss": 0.0476, "step": 40820 }, { "epoch": 0.8159146317093642, "grad_norm": 0.13518232107162476, "learning_rate": 8.892506100737868e-05, "loss": 0.0351, "step": 40830 }, { "epoch": 0.8161144638503657, "grad_norm": 0.11732786148786545, "learning_rate": 8.891857399534654e-05, "loss": 0.0316, "step": 40840 }, { "epoch": 0.8163142959913673, "grad_norm": 0.14301477372646332, "learning_rate": 8.891208532078439e-05, "loss": 0.0401, "step": 40850 }, { "epoch": 0.8165141281323688, "grad_norm": 0.11437970399856567, "learning_rate": 8.890559498396945e-05, "loss": 0.0376, "step": 40860 }, { "epoch": 0.8167139602733704, "grad_norm": 0.14866803586483002, "learning_rate": 8.889910298517896e-05, "loss": 0.0421, "step": 40870 }, { "epoch": 0.8169137924143719, "grad_norm": 0.11676501482725143, "learning_rate": 8.889260932469025e-05, "loss": 0.0401, "step": 40880 }, { "epoch": 0.8171136245553735, "grad_norm": 0.10891705006361008, "learning_rate": 8.88861140027807e-05, "loss": 0.0282, "step": 40890 }, { "epoch": 0.817313456696375, "grad_norm": 0.17681001126766205, "learning_rate": 8.887961701972783e-05, "loss": 0.0379, "step": 40900 }, { "epoch": 0.8175132888373766, "grad_norm": 0.14596135914325714, "learning_rate": 8.887311837580914e-05, "loss": 0.038, "step": 40910 }, { "epoch": 0.8177131209783781, "grad_norm": 0.1337902545928955, "learning_rate": 8.886661807130223e-05, "loss": 0.0332, "step": 40920 }, { "epoch": 0.8179129531193797, "grad_norm": 0.13990293443202972, "learning_rate": 8.886011610648479e-05, "loss": 0.033, "step": 40930 }, { "epoch": 0.8181127852603812, "grad_norm": 0.1335011124610901, "learning_rate": 8.88536124816346e-05, "loss": 0.0374, "step": 40940 }, { "epoch": 0.8183126174013828, "grad_norm": 0.1389094442129135, "learning_rate": 8.884710719702945e-05, "loss": 0.0359, "step": 40950 }, { "epoch": 0.8185124495423844, "grad_norm": 0.14485308527946472, "learning_rate": 8.884060025294725e-05, "loss": 0.0329, "step": 40960 }, { "epoch": 0.818712281683386, "grad_norm": 0.12434705346822739, "learning_rate": 8.883409164966593e-05, "loss": 0.0376, "step": 40970 }, { "epoch": 0.8189121138243876, "grad_norm": 0.13954971730709076, "learning_rate": 8.88275813874636e-05, "loss": 0.0323, "step": 40980 }, { "epoch": 0.819111945965389, "grad_norm": 0.13444051146507263, "learning_rate": 8.88210694666183e-05, "loss": 0.0343, "step": 40990 }, { "epoch": 0.8193117781063907, "grad_norm": 0.14748729765415192, "learning_rate": 8.881455588740824e-05, "loss": 0.0342, "step": 41000 }, { "epoch": 0.8195116102473922, "grad_norm": 0.1519659012556076, "learning_rate": 8.880804065011163e-05, "loss": 0.0443, "step": 41010 }, { "epoch": 0.8197114423883938, "grad_norm": 0.1425234079360962, "learning_rate": 8.880152375500685e-05, "loss": 0.0372, "step": 41020 }, { "epoch": 0.8199112745293953, "grad_norm": 0.1286352276802063, "learning_rate": 8.879500520237224e-05, "loss": 0.0364, "step": 41030 }, { "epoch": 0.8201111066703969, "grad_norm": 0.1291033774614334, "learning_rate": 8.878848499248628e-05, "loss": 0.0413, "step": 41040 }, { "epoch": 0.8203109388113984, "grad_norm": 0.15408723056316376, "learning_rate": 8.878196312562749e-05, "loss": 0.0417, "step": 41050 }, { "epoch": 0.8205107709524, "grad_norm": 0.13778765499591827, "learning_rate": 8.877543960207449e-05, "loss": 0.0337, "step": 41060 }, { "epoch": 0.8207106030934015, "grad_norm": 0.13118745386600494, "learning_rate": 8.876891442210593e-05, "loss": 0.0415, "step": 41070 }, { "epoch": 0.8209104352344031, "grad_norm": 0.14846783876419067, "learning_rate": 8.876238758600057e-05, "loss": 0.0364, "step": 41080 }, { "epoch": 0.8211102673754047, "grad_norm": 0.13947923481464386, "learning_rate": 8.875585909403723e-05, "loss": 0.0319, "step": 41090 }, { "epoch": 0.8213100995164062, "grad_norm": 0.11986514180898666, "learning_rate": 8.874932894649479e-05, "loss": 0.0416, "step": 41100 }, { "epoch": 0.8215099316574078, "grad_norm": 0.18660713732242584, "learning_rate": 8.874279714365219e-05, "loss": 0.0374, "step": 41110 }, { "epoch": 0.8217097637984093, "grad_norm": 0.14226049184799194, "learning_rate": 8.873626368578847e-05, "loss": 0.0424, "step": 41120 }, { "epoch": 0.8219095959394109, "grad_norm": 0.1426001638174057, "learning_rate": 8.872972857318274e-05, "loss": 0.0388, "step": 41130 }, { "epoch": 0.8221094280804124, "grad_norm": 0.12566250562667847, "learning_rate": 8.872319180611412e-05, "loss": 0.0366, "step": 41140 }, { "epoch": 0.822309260221414, "grad_norm": 0.13071441650390625, "learning_rate": 8.871665338486192e-05, "loss": 0.0392, "step": 41150 }, { "epoch": 0.8225090923624155, "grad_norm": 0.13987281918525696, "learning_rate": 8.87101133097054e-05, "loss": 0.0392, "step": 41160 }, { "epoch": 0.8227089245034171, "grad_norm": 0.1128038614988327, "learning_rate": 8.870357158092394e-05, "loss": 0.032, "step": 41170 }, { "epoch": 0.8229087566444186, "grad_norm": 0.11253658682107925, "learning_rate": 8.869702819879701e-05, "loss": 0.0361, "step": 41180 }, { "epoch": 0.8231085887854203, "grad_norm": 0.11517590284347534, "learning_rate": 8.869048316360413e-05, "loss": 0.0391, "step": 41190 }, { "epoch": 0.8233084209264218, "grad_norm": 0.11596968024969101, "learning_rate": 8.86839364756249e-05, "loss": 0.0462, "step": 41200 }, { "epoch": 0.8235082530674234, "grad_norm": 0.13095618784427643, "learning_rate": 8.867738813513894e-05, "loss": 0.0444, "step": 41210 }, { "epoch": 0.823708085208425, "grad_norm": 0.1453198939561844, "learning_rate": 8.867083814242603e-05, "loss": 0.0373, "step": 41220 }, { "epoch": 0.8239079173494265, "grad_norm": 0.16364887356758118, "learning_rate": 8.866428649776594e-05, "loss": 0.0367, "step": 41230 }, { "epoch": 0.8241077494904281, "grad_norm": 0.12353340536355972, "learning_rate": 8.865773320143857e-05, "loss": 0.032, "step": 41240 }, { "epoch": 0.8243075816314296, "grad_norm": 0.14967741072177887, "learning_rate": 8.865117825372385e-05, "loss": 0.0311, "step": 41250 }, { "epoch": 0.8245074137724312, "grad_norm": 0.14899612963199615, "learning_rate": 8.864462165490179e-05, "loss": 0.0372, "step": 41260 }, { "epoch": 0.8247072459134327, "grad_norm": 0.13977782428264618, "learning_rate": 8.863806340525251e-05, "loss": 0.038, "step": 41270 }, { "epoch": 0.8249070780544343, "grad_norm": 0.13319650292396545, "learning_rate": 8.863150350505613e-05, "loss": 0.0323, "step": 41280 }, { "epoch": 0.8251069101954358, "grad_norm": 0.17730732262134552, "learning_rate": 8.862494195459288e-05, "loss": 0.0349, "step": 41290 }, { "epoch": 0.8253067423364374, "grad_norm": 0.1464810073375702, "learning_rate": 8.861837875414308e-05, "loss": 0.0438, "step": 41300 }, { "epoch": 0.8255065744774389, "grad_norm": 0.11557640880346298, "learning_rate": 8.861181390398708e-05, "loss": 0.035, "step": 41310 }, { "epoch": 0.8257064066184405, "grad_norm": 0.10767465829849243, "learning_rate": 8.86052474044053e-05, "loss": 0.0463, "step": 41320 }, { "epoch": 0.825906238759442, "grad_norm": 0.13926272094249725, "learning_rate": 8.85986792556783e-05, "loss": 0.0394, "step": 41330 }, { "epoch": 0.8261060709004436, "grad_norm": 0.1197480633854866, "learning_rate": 8.859210945808662e-05, "loss": 0.0386, "step": 41340 }, { "epoch": 0.8263059030414451, "grad_norm": 0.12423216551542282, "learning_rate": 8.858553801191092e-05, "loss": 0.0336, "step": 41350 }, { "epoch": 0.8265057351824467, "grad_norm": 0.1283731758594513, "learning_rate": 8.857896491743192e-05, "loss": 0.0343, "step": 41360 }, { "epoch": 0.8267055673234484, "grad_norm": 0.18259860575199127, "learning_rate": 8.857239017493042e-05, "loss": 0.0378, "step": 41370 }, { "epoch": 0.8269053994644499, "grad_norm": 0.12817275524139404, "learning_rate": 8.856581378468725e-05, "loss": 0.0396, "step": 41380 }, { "epoch": 0.8271052316054515, "grad_norm": 0.14740711450576782, "learning_rate": 8.855923574698337e-05, "loss": 0.0364, "step": 41390 }, { "epoch": 0.827305063746453, "grad_norm": 0.16482755541801453, "learning_rate": 8.855265606209979e-05, "loss": 0.0384, "step": 41400 }, { "epoch": 0.8275048958874546, "grad_norm": 0.14911824464797974, "learning_rate": 8.854607473031756e-05, "loss": 0.0335, "step": 41410 }, { "epoch": 0.8277047280284561, "grad_norm": 0.1379098892211914, "learning_rate": 8.853949175191782e-05, "loss": 0.0391, "step": 41420 }, { "epoch": 0.8279045601694577, "grad_norm": 0.13186056911945343, "learning_rate": 8.853290712718179e-05, "loss": 0.04, "step": 41430 }, { "epoch": 0.8281043923104592, "grad_norm": 0.13409028947353363, "learning_rate": 8.852632085639077e-05, "loss": 0.0371, "step": 41440 }, { "epoch": 0.8283042244514608, "grad_norm": 0.13894285261631012, "learning_rate": 8.851973293982608e-05, "loss": 0.0362, "step": 41450 }, { "epoch": 0.8285040565924623, "grad_norm": 0.1538274735212326, "learning_rate": 8.851314337776917e-05, "loss": 0.0436, "step": 41460 }, { "epoch": 0.8287038887334639, "grad_norm": 0.1239919513463974, "learning_rate": 8.850655217050154e-05, "loss": 0.032, "step": 41470 }, { "epoch": 0.8289037208744654, "grad_norm": 0.14453770220279694, "learning_rate": 8.849995931830472e-05, "loss": 0.0367, "step": 41480 }, { "epoch": 0.829103553015467, "grad_norm": 0.13424119353294373, "learning_rate": 8.849336482146038e-05, "loss": 0.032, "step": 41490 }, { "epoch": 0.8293033851564686, "grad_norm": 0.1521206647157669, "learning_rate": 8.848676868025018e-05, "loss": 0.0334, "step": 41500 }, { "epoch": 0.8295032172974701, "grad_norm": 0.12230881303548813, "learning_rate": 8.848017089495595e-05, "loss": 0.0288, "step": 41510 }, { "epoch": 0.8297030494384717, "grad_norm": 0.13524499535560608, "learning_rate": 8.84735714658595e-05, "loss": 0.0394, "step": 41520 }, { "epoch": 0.8299028815794732, "grad_norm": 0.1310303807258606, "learning_rate": 8.846697039324276e-05, "loss": 0.0404, "step": 41530 }, { "epoch": 0.8301027137204748, "grad_norm": 0.14268885552883148, "learning_rate": 8.846036767738773e-05, "loss": 0.035, "step": 41540 }, { "epoch": 0.8303025458614763, "grad_norm": 0.16475693881511688, "learning_rate": 8.845376331857643e-05, "loss": 0.0468, "step": 41550 }, { "epoch": 0.830502378002478, "grad_norm": 0.14424972236156464, "learning_rate": 8.844715731709098e-05, "loss": 0.0391, "step": 41560 }, { "epoch": 0.8307022101434794, "grad_norm": 0.2088807225227356, "learning_rate": 8.844054967321363e-05, "loss": 0.0392, "step": 41570 }, { "epoch": 0.8309020422844811, "grad_norm": 0.1749555766582489, "learning_rate": 8.843394038722661e-05, "loss": 0.0421, "step": 41580 }, { "epoch": 0.8311018744254826, "grad_norm": 0.12600086629390717, "learning_rate": 8.842732945941225e-05, "loss": 0.0404, "step": 41590 }, { "epoch": 0.8313017065664842, "grad_norm": 0.1351122260093689, "learning_rate": 8.842071689005297e-05, "loss": 0.0384, "step": 41600 }, { "epoch": 0.8315015387074857, "grad_norm": 0.1119481697678566, "learning_rate": 8.841410267943127e-05, "loss": 0.0323, "step": 41610 }, { "epoch": 0.8317013708484873, "grad_norm": 0.13713973760604858, "learning_rate": 8.840748682782964e-05, "loss": 0.0426, "step": 41620 }, { "epoch": 0.8319012029894888, "grad_norm": 0.13414283096790314, "learning_rate": 8.840086933553075e-05, "loss": 0.0373, "step": 41630 }, { "epoch": 0.8321010351304904, "grad_norm": 0.17145627737045288, "learning_rate": 8.839425020281727e-05, "loss": 0.0371, "step": 41640 }, { "epoch": 0.832300867271492, "grad_norm": 0.13573357462882996, "learning_rate": 8.838762942997194e-05, "loss": 0.0348, "step": 41650 }, { "epoch": 0.8325006994124935, "grad_norm": 0.16587458550930023, "learning_rate": 8.838100701727761e-05, "loss": 0.0398, "step": 41660 }, { "epoch": 0.8327005315534951, "grad_norm": 0.1370893120765686, "learning_rate": 8.837438296501714e-05, "loss": 0.0332, "step": 41670 }, { "epoch": 0.8329003636944966, "grad_norm": 0.13533666729927063, "learning_rate": 8.836775727347356e-05, "loss": 0.0438, "step": 41680 }, { "epoch": 0.8331001958354982, "grad_norm": 0.16684463620185852, "learning_rate": 8.836112994292986e-05, "loss": 0.0448, "step": 41690 }, { "epoch": 0.8333000279764997, "grad_norm": 0.17788399755954742, "learning_rate": 8.835450097366918e-05, "loss": 0.0336, "step": 41700 }, { "epoch": 0.8334998601175013, "grad_norm": 0.15156720578670502, "learning_rate": 8.834787036597465e-05, "loss": 0.0362, "step": 41710 }, { "epoch": 0.8336996922585028, "grad_norm": 0.15656062960624695, "learning_rate": 8.834123812012956e-05, "loss": 0.0271, "step": 41720 }, { "epoch": 0.8338995243995044, "grad_norm": 0.24623189866542816, "learning_rate": 8.833460423641719e-05, "loss": 0.0347, "step": 41730 }, { "epoch": 0.8340993565405059, "grad_norm": 0.1280488669872284, "learning_rate": 8.832796871512097e-05, "loss": 0.0338, "step": 41740 }, { "epoch": 0.8342991886815075, "grad_norm": 0.1003110408782959, "learning_rate": 8.832133155652435e-05, "loss": 0.0328, "step": 41750 }, { "epoch": 0.834499020822509, "grad_norm": 0.12557995319366455, "learning_rate": 8.831469276091082e-05, "loss": 0.0389, "step": 41760 }, { "epoch": 0.8346988529635107, "grad_norm": 0.18888182938098907, "learning_rate": 8.830805232856402e-05, "loss": 0.0309, "step": 41770 }, { "epoch": 0.8348986851045123, "grad_norm": 0.1458241194486618, "learning_rate": 8.830141025976759e-05, "loss": 0.0367, "step": 41780 }, { "epoch": 0.8350985172455138, "grad_norm": 0.1290023922920227, "learning_rate": 8.829476655480529e-05, "loss": 0.042, "step": 41790 }, { "epoch": 0.8352983493865154, "grad_norm": 0.1431794911623001, "learning_rate": 8.828812121396089e-05, "loss": 0.0336, "step": 41800 }, { "epoch": 0.8354981815275169, "grad_norm": 0.16368192434310913, "learning_rate": 8.828147423751829e-05, "loss": 0.0404, "step": 41810 }, { "epoch": 0.8356980136685185, "grad_norm": 0.1319800615310669, "learning_rate": 8.827482562576143e-05, "loss": 0.0342, "step": 41820 }, { "epoch": 0.83589784580952, "grad_norm": 0.13944154977798462, "learning_rate": 8.826817537897437e-05, "loss": 0.0379, "step": 41830 }, { "epoch": 0.8360976779505216, "grad_norm": 0.17361587285995483, "learning_rate": 8.826152349744113e-05, "loss": 0.0456, "step": 41840 }, { "epoch": 0.8362975100915231, "grad_norm": 0.13323670625686646, "learning_rate": 8.825486998144589e-05, "loss": 0.0414, "step": 41850 }, { "epoch": 0.8364973422325247, "grad_norm": 0.21555683016777039, "learning_rate": 8.82482148312729e-05, "loss": 0.0284, "step": 41860 }, { "epoch": 0.8366971743735262, "grad_norm": 0.12970241904258728, "learning_rate": 8.824155804720642e-05, "loss": 0.0326, "step": 41870 }, { "epoch": 0.8368970065145278, "grad_norm": 0.1552884727716446, "learning_rate": 8.823489962953085e-05, "loss": 0.0398, "step": 41880 }, { "epoch": 0.8370968386555293, "grad_norm": 0.13353805243968964, "learning_rate": 8.82282395785306e-05, "loss": 0.0348, "step": 41890 }, { "epoch": 0.8372966707965309, "grad_norm": 0.1436821073293686, "learning_rate": 8.822157789449018e-05, "loss": 0.0328, "step": 41900 }, { "epoch": 0.8374965029375325, "grad_norm": 0.14903631806373596, "learning_rate": 8.821491457769416e-05, "loss": 0.0411, "step": 41910 }, { "epoch": 0.837696335078534, "grad_norm": 0.12328886985778809, "learning_rate": 8.82082496284272e-05, "loss": 0.0361, "step": 41920 }, { "epoch": 0.8378961672195356, "grad_norm": 0.17292797565460205, "learning_rate": 8.820158304697401e-05, "loss": 0.0461, "step": 41930 }, { "epoch": 0.8380959993605371, "grad_norm": 0.12666617333889008, "learning_rate": 8.819491483361937e-05, "loss": 0.0363, "step": 41940 }, { "epoch": 0.8382958315015387, "grad_norm": 0.18119721114635468, "learning_rate": 8.818824498864813e-05, "loss": 0.0402, "step": 41950 }, { "epoch": 0.8384956636425402, "grad_norm": 0.12092489749193192, "learning_rate": 8.818157351234521e-05, "loss": 0.0358, "step": 41960 }, { "epoch": 0.8386954957835419, "grad_norm": 0.11327093839645386, "learning_rate": 8.817490040499563e-05, "loss": 0.0452, "step": 41970 }, { "epoch": 0.8388953279245434, "grad_norm": 0.13983705639839172, "learning_rate": 8.816822566688442e-05, "loss": 0.0357, "step": 41980 }, { "epoch": 0.839095160065545, "grad_norm": 0.12404042482376099, "learning_rate": 8.816154929829675e-05, "loss": 0.0437, "step": 41990 }, { "epoch": 0.8392949922065465, "grad_norm": 0.14390243589878082, "learning_rate": 8.815487129951776e-05, "loss": 0.0426, "step": 42000 }, { "epoch": 0.8394948243475481, "grad_norm": 0.12457159161567688, "learning_rate": 8.81481916708328e-05, "loss": 0.0315, "step": 42010 }, { "epoch": 0.8396946564885496, "grad_norm": 0.10194064676761627, "learning_rate": 8.814151041252714e-05, "loss": 0.0364, "step": 42020 }, { "epoch": 0.8398944886295512, "grad_norm": 0.14208398759365082, "learning_rate": 8.813482752488624e-05, "loss": 0.0345, "step": 42030 }, { "epoch": 0.8400943207705527, "grad_norm": 0.1293412744998932, "learning_rate": 8.812814300819556e-05, "loss": 0.0321, "step": 42040 }, { "epoch": 0.8402941529115543, "grad_norm": 0.1635836660861969, "learning_rate": 8.812145686274066e-05, "loss": 0.0404, "step": 42050 }, { "epoch": 0.8404939850525559, "grad_norm": 0.1198403611779213, "learning_rate": 8.811476908880714e-05, "loss": 0.0327, "step": 42060 }, { "epoch": 0.8406938171935574, "grad_norm": 0.13762357831001282, "learning_rate": 8.810807968668072e-05, "loss": 0.0324, "step": 42070 }, { "epoch": 0.840893649334559, "grad_norm": 0.1543557047843933, "learning_rate": 8.810138865664714e-05, "loss": 0.0383, "step": 42080 }, { "epoch": 0.8410934814755605, "grad_norm": 0.13371670246124268, "learning_rate": 8.809469599899223e-05, "loss": 0.0406, "step": 42090 }, { "epoch": 0.8412933136165621, "grad_norm": 0.14874009788036346, "learning_rate": 8.80880017140019e-05, "loss": 0.0395, "step": 42100 }, { "epoch": 0.8414931457575636, "grad_norm": 0.13313190639019012, "learning_rate": 8.808130580196208e-05, "loss": 0.0385, "step": 42110 }, { "epoch": 0.8416929778985652, "grad_norm": 0.11549793928861618, "learning_rate": 8.807460826315885e-05, "loss": 0.0394, "step": 42120 }, { "epoch": 0.8418928100395667, "grad_norm": 0.1157296672463417, "learning_rate": 8.806790909787828e-05, "loss": 0.0324, "step": 42130 }, { "epoch": 0.8420926421805683, "grad_norm": 0.13876795768737793, "learning_rate": 8.80612083064066e-05, "loss": 0.0368, "step": 42140 }, { "epoch": 0.8422924743215698, "grad_norm": 0.2023402750492096, "learning_rate": 8.805450588902999e-05, "loss": 0.0448, "step": 42150 }, { "epoch": 0.8424923064625714, "grad_norm": 0.16001607477664948, "learning_rate": 8.804780184603482e-05, "loss": 0.0411, "step": 42160 }, { "epoch": 0.842692138603573, "grad_norm": 0.17042182385921478, "learning_rate": 8.804109617770743e-05, "loss": 0.0386, "step": 42170 }, { "epoch": 0.8428919707445746, "grad_norm": 0.1360424906015396, "learning_rate": 8.803438888433433e-05, "loss": 0.0312, "step": 42180 }, { "epoch": 0.8430918028855762, "grad_norm": 0.12080464512109756, "learning_rate": 8.802767996620199e-05, "loss": 0.0396, "step": 42190 }, { "epoch": 0.8432916350265777, "grad_norm": 0.1159607470035553, "learning_rate": 8.802096942359701e-05, "loss": 0.0431, "step": 42200 }, { "epoch": 0.8434914671675793, "grad_norm": 0.13304947316646576, "learning_rate": 8.801425725680608e-05, "loss": 0.0371, "step": 42210 }, { "epoch": 0.8436912993085808, "grad_norm": 0.12149479240179062, "learning_rate": 8.80075434661159e-05, "loss": 0.0354, "step": 42220 }, { "epoch": 0.8438911314495824, "grad_norm": 0.14270834624767303, "learning_rate": 8.800082805181329e-05, "loss": 0.0419, "step": 42230 }, { "epoch": 0.8440909635905839, "grad_norm": 0.11562980711460114, "learning_rate": 8.79941110141851e-05, "loss": 0.0332, "step": 42240 }, { "epoch": 0.8442907957315855, "grad_norm": 0.10350947827100754, "learning_rate": 8.798739235351831e-05, "loss": 0.0306, "step": 42250 }, { "epoch": 0.844490627872587, "grad_norm": 0.12345825135707855, "learning_rate": 8.79806720700999e-05, "loss": 0.0352, "step": 42260 }, { "epoch": 0.8446904600135886, "grad_norm": 0.12079767137765884, "learning_rate": 8.797395016421694e-05, "loss": 0.0413, "step": 42270 }, { "epoch": 0.8448902921545901, "grad_norm": 0.1639714390039444, "learning_rate": 8.79672266361566e-05, "loss": 0.0362, "step": 42280 }, { "epoch": 0.8450901242955917, "grad_norm": 0.13603800535202026, "learning_rate": 8.796050148620607e-05, "loss": 0.0386, "step": 42290 }, { "epoch": 0.8452899564365932, "grad_norm": 0.11670761555433273, "learning_rate": 8.795377471465268e-05, "loss": 0.0372, "step": 42300 }, { "epoch": 0.8454897885775948, "grad_norm": 0.13050802052021027, "learning_rate": 8.794704632178373e-05, "loss": 0.0396, "step": 42310 }, { "epoch": 0.8456896207185963, "grad_norm": 0.11765624582767487, "learning_rate": 8.794031630788668e-05, "loss": 0.0301, "step": 42320 }, { "epoch": 0.8458894528595979, "grad_norm": 0.14941725134849548, "learning_rate": 8.793358467324902e-05, "loss": 0.0378, "step": 42330 }, { "epoch": 0.8460892850005995, "grad_norm": 0.10422749817371368, "learning_rate": 8.79268514181583e-05, "loss": 0.0401, "step": 42340 }, { "epoch": 0.846289117141601, "grad_norm": 0.13965758681297302, "learning_rate": 8.792011654290219e-05, "loss": 0.0329, "step": 42350 }, { "epoch": 0.8464889492826027, "grad_norm": 0.13632792234420776, "learning_rate": 8.791338004776833e-05, "loss": 0.0356, "step": 42360 }, { "epoch": 0.8466887814236042, "grad_norm": 0.1175181195139885, "learning_rate": 8.790664193304454e-05, "loss": 0.0284, "step": 42370 }, { "epoch": 0.8468886135646058, "grad_norm": 0.21609176695346832, "learning_rate": 8.789990219901865e-05, "loss": 0.0378, "step": 42380 }, { "epoch": 0.8470884457056073, "grad_norm": 0.12837329506874084, "learning_rate": 8.789316084597854e-05, "loss": 0.0415, "step": 42390 }, { "epoch": 0.8472882778466089, "grad_norm": 0.1519363820552826, "learning_rate": 8.788641787421223e-05, "loss": 0.0342, "step": 42400 }, { "epoch": 0.8474881099876104, "grad_norm": 0.12496566772460938, "learning_rate": 8.787967328400773e-05, "loss": 0.0359, "step": 42410 }, { "epoch": 0.847687942128612, "grad_norm": 0.17504379153251648, "learning_rate": 8.787292707565319e-05, "loss": 0.0324, "step": 42420 }, { "epoch": 0.8478877742696135, "grad_norm": 0.13590484857559204, "learning_rate": 8.786617924943677e-05, "loss": 0.0394, "step": 42430 }, { "epoch": 0.8480876064106151, "grad_norm": 0.12050874531269073, "learning_rate": 8.785942980564675e-05, "loss": 0.033, "step": 42440 }, { "epoch": 0.8482874385516166, "grad_norm": 0.14148260653018951, "learning_rate": 8.785267874457141e-05, "loss": 0.0412, "step": 42450 }, { "epoch": 0.8484872706926182, "grad_norm": 0.12075044959783554, "learning_rate": 8.78459260664992e-05, "loss": 0.0395, "step": 42460 }, { "epoch": 0.8486871028336198, "grad_norm": 0.14163576066493988, "learning_rate": 8.783917177171854e-05, "loss": 0.0421, "step": 42470 }, { "epoch": 0.8488869349746213, "grad_norm": 0.13860082626342773, "learning_rate": 8.783241586051799e-05, "loss": 0.0368, "step": 42480 }, { "epoch": 0.8490867671156229, "grad_norm": 0.15406093001365662, "learning_rate": 8.782565833318612e-05, "loss": 0.038, "step": 42490 }, { "epoch": 0.8492865992566244, "grad_norm": 0.13032153248786926, "learning_rate": 8.781889919001161e-05, "loss": 0.0356, "step": 42500 }, { "epoch": 0.849486431397626, "grad_norm": 0.13300316035747528, "learning_rate": 8.781213843128322e-05, "loss": 0.0382, "step": 42510 }, { "epoch": 0.8496862635386275, "grad_norm": 0.14023061096668243, "learning_rate": 8.780537605728972e-05, "loss": 0.0406, "step": 42520 }, { "epoch": 0.8498860956796291, "grad_norm": 0.17661692202091217, "learning_rate": 8.779861206832e-05, "loss": 0.044, "step": 42530 }, { "epoch": 0.8500859278206306, "grad_norm": 0.12060722708702087, "learning_rate": 8.779184646466304e-05, "loss": 0.0379, "step": 42540 }, { "epoch": 0.8502857599616322, "grad_norm": 0.1429709941148758, "learning_rate": 8.778507924660781e-05, "loss": 0.0456, "step": 42550 }, { "epoch": 0.8504855921026337, "grad_norm": 0.10461752116680145, "learning_rate": 8.777831041444342e-05, "loss": 0.0349, "step": 42560 }, { "epoch": 0.8506854242436354, "grad_norm": 0.16144080460071564, "learning_rate": 8.777153996845898e-05, "loss": 0.0313, "step": 42570 }, { "epoch": 0.8508852563846369, "grad_norm": 0.17834456264972687, "learning_rate": 8.776476790894377e-05, "loss": 0.0409, "step": 42580 }, { "epoch": 0.8510850885256385, "grad_norm": 0.1336652785539627, "learning_rate": 8.775799423618704e-05, "loss": 0.044, "step": 42590 }, { "epoch": 0.8512849206666401, "grad_norm": 0.11160160601139069, "learning_rate": 8.775121895047816e-05, "loss": 0.041, "step": 42600 }, { "epoch": 0.8514847528076416, "grad_norm": 0.1421736776828766, "learning_rate": 8.774444205210654e-05, "loss": 0.0257, "step": 42610 }, { "epoch": 0.8516845849486432, "grad_norm": 0.13089774549007416, "learning_rate": 8.773766354136173e-05, "loss": 0.0348, "step": 42620 }, { "epoch": 0.8518844170896447, "grad_norm": 0.11793188005685806, "learning_rate": 8.773088341853323e-05, "loss": 0.0326, "step": 42630 }, { "epoch": 0.8520842492306463, "grad_norm": 0.14533023536205292, "learning_rate": 8.772410168391072e-05, "loss": 0.0359, "step": 42640 }, { "epoch": 0.8522840813716478, "grad_norm": 0.10905122011899948, "learning_rate": 8.771731833778388e-05, "loss": 0.0382, "step": 42650 }, { "epoch": 0.8524839135126494, "grad_norm": 0.15902285277843475, "learning_rate": 8.77105333804425e-05, "loss": 0.0331, "step": 42660 }, { "epoch": 0.8526837456536509, "grad_norm": 0.12875501811504364, "learning_rate": 8.77037468121764e-05, "loss": 0.0389, "step": 42670 }, { "epoch": 0.8528835777946525, "grad_norm": 0.17643824219703674, "learning_rate": 8.769695863327551e-05, "loss": 0.0365, "step": 42680 }, { "epoch": 0.853083409935654, "grad_norm": 0.29587915539741516, "learning_rate": 8.769016884402979e-05, "loss": 0.0335, "step": 42690 }, { "epoch": 0.8532832420766556, "grad_norm": 0.11160390079021454, "learning_rate": 8.76833774447293e-05, "loss": 0.036, "step": 42700 }, { "epoch": 0.8534830742176571, "grad_norm": 0.15384072065353394, "learning_rate": 8.767658443566415e-05, "loss": 0.0431, "step": 42710 }, { "epoch": 0.8536829063586587, "grad_norm": 0.15211644768714905, "learning_rate": 8.766978981712452e-05, "loss": 0.0436, "step": 42720 }, { "epoch": 0.8538827384996602, "grad_norm": 0.15671493113040924, "learning_rate": 8.766299358940067e-05, "loss": 0.0429, "step": 42730 }, { "epoch": 0.8540825706406618, "grad_norm": 0.19804732501506805, "learning_rate": 8.765619575278295e-05, "loss": 0.0376, "step": 42740 }, { "epoch": 0.8542824027816635, "grad_norm": 0.16757117211818695, "learning_rate": 8.764939630756169e-05, "loss": 0.0379, "step": 42750 }, { "epoch": 0.854482234922665, "grad_norm": 0.15216417610645294, "learning_rate": 8.764259525402741e-05, "loss": 0.0401, "step": 42760 }, { "epoch": 0.8546820670636666, "grad_norm": 0.1406945139169693, "learning_rate": 8.763579259247059e-05, "loss": 0.0354, "step": 42770 }, { "epoch": 0.8548818992046681, "grad_norm": 0.1495646983385086, "learning_rate": 8.762898832318187e-05, "loss": 0.0356, "step": 42780 }, { "epoch": 0.8550817313456697, "grad_norm": 0.13272710144519806, "learning_rate": 8.762218244645189e-05, "loss": 0.0462, "step": 42790 }, { "epoch": 0.8552815634866712, "grad_norm": 0.12574777007102966, "learning_rate": 8.761537496257138e-05, "loss": 0.0331, "step": 42800 }, { "epoch": 0.8554813956276728, "grad_norm": 0.14258041977882385, "learning_rate": 8.760856587183116e-05, "loss": 0.0367, "step": 42810 }, { "epoch": 0.8556812277686743, "grad_norm": 0.13313807547092438, "learning_rate": 8.76017551745221e-05, "loss": 0.0377, "step": 42820 }, { "epoch": 0.8558810599096759, "grad_norm": 0.12165360152721405, "learning_rate": 8.759494287093512e-05, "loss": 0.0455, "step": 42830 }, { "epoch": 0.8560808920506774, "grad_norm": 0.16542822122573853, "learning_rate": 8.758812896136126e-05, "loss": 0.0328, "step": 42840 }, { "epoch": 0.856280724191679, "grad_norm": 0.13040336966514587, "learning_rate": 8.758131344609158e-05, "loss": 0.042, "step": 42850 }, { "epoch": 0.8564805563326805, "grad_norm": 0.15657368302345276, "learning_rate": 8.757449632541722e-05, "loss": 0.0392, "step": 42860 }, { "epoch": 0.8566803884736821, "grad_norm": 0.15017031133174896, "learning_rate": 8.756767759962941e-05, "loss": 0.0371, "step": 42870 }, { "epoch": 0.8568802206146837, "grad_norm": 0.14644213020801544, "learning_rate": 8.756085726901943e-05, "loss": 0.0401, "step": 42880 }, { "epoch": 0.8570800527556852, "grad_norm": 0.14310789108276367, "learning_rate": 8.755403533387863e-05, "loss": 0.041, "step": 42890 }, { "epoch": 0.8572798848966868, "grad_norm": 0.16400010883808136, "learning_rate": 8.754721179449842e-05, "loss": 0.0403, "step": 42900 }, { "epoch": 0.8574797170376883, "grad_norm": 0.12256735563278198, "learning_rate": 8.75403866511703e-05, "loss": 0.0367, "step": 42910 }, { "epoch": 0.8576795491786899, "grad_norm": 0.1495232880115509, "learning_rate": 8.753355990418583e-05, "loss": 0.0364, "step": 42920 }, { "epoch": 0.8578793813196914, "grad_norm": 0.11043941974639893, "learning_rate": 8.752673155383662e-05, "loss": 0.0353, "step": 42930 }, { "epoch": 0.858079213460693, "grad_norm": 0.17988847196102142, "learning_rate": 8.751990160041439e-05, "loss": 0.0442, "step": 42940 }, { "epoch": 0.8582790456016945, "grad_norm": 0.1354890614748001, "learning_rate": 8.75130700442109e-05, "loss": 0.0301, "step": 42950 }, { "epoch": 0.8584788777426962, "grad_norm": 0.11692377179861069, "learning_rate": 8.750623688551795e-05, "loss": 0.038, "step": 42960 }, { "epoch": 0.8586787098836977, "grad_norm": 0.18596184253692627, "learning_rate": 8.749940212462748e-05, "loss": 0.0377, "step": 42970 }, { "epoch": 0.8588785420246993, "grad_norm": 0.1448715776205063, "learning_rate": 8.749256576183142e-05, "loss": 0.0396, "step": 42980 }, { "epoch": 0.8590783741657008, "grad_norm": 0.1283327043056488, "learning_rate": 8.748572779742185e-05, "loss": 0.0386, "step": 42990 }, { "epoch": 0.8592782063067024, "grad_norm": 0.11236979812383652, "learning_rate": 8.747888823169084e-05, "loss": 0.0331, "step": 43000 }, { "epoch": 0.8594780384477039, "grad_norm": 0.16440081596374512, "learning_rate": 8.747204706493057e-05, "loss": 0.0377, "step": 43010 }, { "epoch": 0.8596778705887055, "grad_norm": 0.20554181933403015, "learning_rate": 8.74652042974333e-05, "loss": 0.0349, "step": 43020 }, { "epoch": 0.8598777027297071, "grad_norm": 0.17906765639781952, "learning_rate": 8.745835992949132e-05, "loss": 0.0383, "step": 43030 }, { "epoch": 0.8600775348707086, "grad_norm": 0.1470031589269638, "learning_rate": 8.745151396139701e-05, "loss": 0.0375, "step": 43040 }, { "epoch": 0.8602773670117102, "grad_norm": 0.1268109530210495, "learning_rate": 8.744466639344284e-05, "loss": 0.032, "step": 43050 }, { "epoch": 0.8604771991527117, "grad_norm": 0.12681308388710022, "learning_rate": 8.74378172259213e-05, "loss": 0.0346, "step": 43060 }, { "epoch": 0.8606770312937133, "grad_norm": 0.12044690549373627, "learning_rate": 8.743096645912499e-05, "loss": 0.0373, "step": 43070 }, { "epoch": 0.8608768634347148, "grad_norm": 0.12539741396903992, "learning_rate": 8.742411409334654e-05, "loss": 0.0332, "step": 43080 }, { "epoch": 0.8610766955757164, "grad_norm": 0.10443530976772308, "learning_rate": 8.74172601288787e-05, "loss": 0.0275, "step": 43090 }, { "epoch": 0.8612765277167179, "grad_norm": 0.11997809261083603, "learning_rate": 8.741040456601425e-05, "loss": 0.0361, "step": 43100 }, { "epoch": 0.8614763598577195, "grad_norm": 0.13398264348506927, "learning_rate": 8.740354740504603e-05, "loss": 0.0332, "step": 43110 }, { "epoch": 0.861676191998721, "grad_norm": 0.16220971941947937, "learning_rate": 8.739668864626699e-05, "loss": 0.0316, "step": 43120 }, { "epoch": 0.8618760241397226, "grad_norm": 0.14375878870487213, "learning_rate": 8.73898282899701e-05, "loss": 0.0405, "step": 43130 }, { "epoch": 0.8620758562807241, "grad_norm": 0.15003517270088196, "learning_rate": 8.738296633644845e-05, "loss": 0.0308, "step": 43140 }, { "epoch": 0.8622756884217258, "grad_norm": 0.15437038242816925, "learning_rate": 8.737610278599513e-05, "loss": 0.039, "step": 43150 }, { "epoch": 0.8624755205627274, "grad_norm": 0.13647347688674927, "learning_rate": 8.736923763890337e-05, "loss": 0.0323, "step": 43160 }, { "epoch": 0.8626753527037289, "grad_norm": 0.170035257935524, "learning_rate": 8.736237089546643e-05, "loss": 0.044, "step": 43170 }, { "epoch": 0.8628751848447305, "grad_norm": 0.14175613224506378, "learning_rate": 8.735550255597765e-05, "loss": 0.0438, "step": 43180 }, { "epoch": 0.863075016985732, "grad_norm": 0.16106191277503967, "learning_rate": 8.73486326207304e-05, "loss": 0.0364, "step": 43190 }, { "epoch": 0.8632748491267336, "grad_norm": 0.11463037878274918, "learning_rate": 8.734176109001819e-05, "loss": 0.0323, "step": 43200 }, { "epoch": 0.8634746812677351, "grad_norm": 0.15470227599143982, "learning_rate": 8.733488796413456e-05, "loss": 0.0382, "step": 43210 }, { "epoch": 0.8636745134087367, "grad_norm": 0.12611636519432068, "learning_rate": 8.732801324337307e-05, "loss": 0.0447, "step": 43220 }, { "epoch": 0.8638743455497382, "grad_norm": 0.109749436378479, "learning_rate": 8.732113692802746e-05, "loss": 0.0392, "step": 43230 }, { "epoch": 0.8640741776907398, "grad_norm": 0.14754478633403778, "learning_rate": 8.731425901839142e-05, "loss": 0.0349, "step": 43240 }, { "epoch": 0.8642740098317413, "grad_norm": 0.10467419773340225, "learning_rate": 8.73073795147588e-05, "loss": 0.032, "step": 43250 }, { "epoch": 0.8644738419727429, "grad_norm": 0.1562776267528534, "learning_rate": 8.730049841742344e-05, "loss": 0.0305, "step": 43260 }, { "epoch": 0.8646736741137444, "grad_norm": 0.1611069142818451, "learning_rate": 8.729361572667934e-05, "loss": 0.0362, "step": 43270 }, { "epoch": 0.864873506254746, "grad_norm": 0.16474202275276184, "learning_rate": 8.728673144282047e-05, "loss": 0.0396, "step": 43280 }, { "epoch": 0.8650733383957476, "grad_norm": 0.14944858849048615, "learning_rate": 8.727984556614093e-05, "loss": 0.0315, "step": 43290 }, { "epoch": 0.8652731705367491, "grad_norm": 0.12641531229019165, "learning_rate": 8.727295809693488e-05, "loss": 0.0381, "step": 43300 }, { "epoch": 0.8654730026777507, "grad_norm": 0.11632051318883896, "learning_rate": 8.726606903549652e-05, "loss": 0.031, "step": 43310 }, { "epoch": 0.8656728348187522, "grad_norm": 0.15256942808628082, "learning_rate": 8.725917838212016e-05, "loss": 0.0357, "step": 43320 }, { "epoch": 0.8658726669597538, "grad_norm": 0.13481496274471283, "learning_rate": 8.725228613710015e-05, "loss": 0.0333, "step": 43330 }, { "epoch": 0.8660724991007553, "grad_norm": 0.14712058007717133, "learning_rate": 8.72453923007309e-05, "loss": 0.0426, "step": 43340 }, { "epoch": 0.866272331241757, "grad_norm": 0.11688540875911713, "learning_rate": 8.723849687330692e-05, "loss": 0.0389, "step": 43350 }, { "epoch": 0.8664721633827585, "grad_norm": 0.11542130261659622, "learning_rate": 8.723159985512277e-05, "loss": 0.0346, "step": 43360 }, { "epoch": 0.8666719955237601, "grad_norm": 0.12855663895606995, "learning_rate": 8.722470124647306e-05, "loss": 0.0475, "step": 43370 }, { "epoch": 0.8668718276647616, "grad_norm": 0.1470254808664322, "learning_rate": 8.721780104765249e-05, "loss": 0.0311, "step": 43380 }, { "epoch": 0.8670716598057632, "grad_norm": 0.1583949625492096, "learning_rate": 8.721089925895585e-05, "loss": 0.0412, "step": 43390 }, { "epoch": 0.8672714919467647, "grad_norm": 0.11979768425226212, "learning_rate": 8.720399588067793e-05, "loss": 0.04, "step": 43400 }, { "epoch": 0.8674713240877663, "grad_norm": 0.12426916509866714, "learning_rate": 8.719709091311368e-05, "loss": 0.0349, "step": 43410 }, { "epoch": 0.8676711562287678, "grad_norm": 0.15278279781341553, "learning_rate": 8.719018435655801e-05, "loss": 0.032, "step": 43420 }, { "epoch": 0.8678709883697694, "grad_norm": 0.14067532122135162, "learning_rate": 8.7183276211306e-05, "loss": 0.0438, "step": 43430 }, { "epoch": 0.868070820510771, "grad_norm": 0.15132582187652588, "learning_rate": 8.717636647765274e-05, "loss": 0.033, "step": 43440 }, { "epoch": 0.8682706526517725, "grad_norm": 0.14582103490829468, "learning_rate": 8.71694551558934e-05, "loss": 0.0342, "step": 43450 }, { "epoch": 0.8684704847927741, "grad_norm": 0.14824466407299042, "learning_rate": 8.71625422463232e-05, "loss": 0.0417, "step": 43460 }, { "epoch": 0.8686703169337756, "grad_norm": 0.12379299104213715, "learning_rate": 8.71556277492375e-05, "loss": 0.0324, "step": 43470 }, { "epoch": 0.8688701490747772, "grad_norm": 0.16814813017845154, "learning_rate": 8.714871166493162e-05, "loss": 0.0306, "step": 43480 }, { "epoch": 0.8690699812157787, "grad_norm": 0.13021980226039886, "learning_rate": 8.714179399370101e-05, "loss": 0.0339, "step": 43490 }, { "epoch": 0.8692698133567803, "grad_norm": 0.13327571749687195, "learning_rate": 8.713487473584121e-05, "loss": 0.0382, "step": 43500 }, { "epoch": 0.8694696454977818, "grad_norm": 0.135309636592865, "learning_rate": 8.712795389164778e-05, "loss": 0.0379, "step": 43510 }, { "epoch": 0.8696694776387834, "grad_norm": 0.15471425652503967, "learning_rate": 8.712103146141635e-05, "loss": 0.035, "step": 43520 }, { "epoch": 0.8698693097797849, "grad_norm": 0.15838859975337982, "learning_rate": 8.711410744544266e-05, "loss": 0.0391, "step": 43530 }, { "epoch": 0.8700691419207865, "grad_norm": 0.13600316643714905, "learning_rate": 8.710718184402248e-05, "loss": 0.0295, "step": 43540 }, { "epoch": 0.870268974061788, "grad_norm": 0.13633322715759277, "learning_rate": 8.710025465745166e-05, "loss": 0.0309, "step": 43550 }, { "epoch": 0.8704688062027897, "grad_norm": 0.1457156240940094, "learning_rate": 8.709332588602612e-05, "loss": 0.0408, "step": 43560 }, { "epoch": 0.8706686383437913, "grad_norm": 0.20911258459091187, "learning_rate": 8.708639553004184e-05, "loss": 0.0386, "step": 43570 }, { "epoch": 0.8708684704847928, "grad_norm": 0.140419140458107, "learning_rate": 8.707946358979486e-05, "loss": 0.0409, "step": 43580 }, { "epoch": 0.8710683026257944, "grad_norm": 0.11850006878376007, "learning_rate": 8.707253006558131e-05, "loss": 0.0308, "step": 43590 }, { "epoch": 0.8712681347667959, "grad_norm": 0.13081876933574677, "learning_rate": 8.70655949576974e-05, "loss": 0.036, "step": 43600 }, { "epoch": 0.8714679669077975, "grad_norm": 0.13385072350502014, "learning_rate": 8.705865826643937e-05, "loss": 0.0341, "step": 43610 }, { "epoch": 0.871667799048799, "grad_norm": 0.1467619240283966, "learning_rate": 8.705171999210353e-05, "loss": 0.0466, "step": 43620 }, { "epoch": 0.8718676311898006, "grad_norm": 0.10722838342189789, "learning_rate": 8.704478013498626e-05, "loss": 0.0351, "step": 43630 }, { "epoch": 0.8720674633308021, "grad_norm": 0.15316802263259888, "learning_rate": 8.703783869538404e-05, "loss": 0.0306, "step": 43640 }, { "epoch": 0.8722672954718037, "grad_norm": 0.12181100994348526, "learning_rate": 8.703089567359341e-05, "loss": 0.0416, "step": 43650 }, { "epoch": 0.8724671276128052, "grad_norm": 0.14864414930343628, "learning_rate": 8.702395106991093e-05, "loss": 0.042, "step": 43660 }, { "epoch": 0.8726669597538068, "grad_norm": 0.11510568112134933, "learning_rate": 8.701700488463329e-05, "loss": 0.0296, "step": 43670 }, { "epoch": 0.8728667918948083, "grad_norm": 0.15591658651828766, "learning_rate": 8.701005711805719e-05, "loss": 0.0384, "step": 43680 }, { "epoch": 0.8730666240358099, "grad_norm": 0.13624195754528046, "learning_rate": 8.700310777047944e-05, "loss": 0.0409, "step": 43690 }, { "epoch": 0.8732664561768114, "grad_norm": 0.1661277562379837, "learning_rate": 8.699615684219691e-05, "loss": 0.044, "step": 43700 }, { "epoch": 0.873466288317813, "grad_norm": 0.16326896846294403, "learning_rate": 8.698920433350653e-05, "loss": 0.0345, "step": 43710 }, { "epoch": 0.8736661204588146, "grad_norm": 0.1274867057800293, "learning_rate": 8.698225024470528e-05, "loss": 0.0374, "step": 43720 }, { "epoch": 0.8738659525998161, "grad_norm": 0.1316165030002594, "learning_rate": 8.697529457609023e-05, "loss": 0.0384, "step": 43730 }, { "epoch": 0.8740657847408178, "grad_norm": 0.1417897641658783, "learning_rate": 8.696833732795853e-05, "loss": 0.0384, "step": 43740 }, { "epoch": 0.8742656168818193, "grad_norm": 0.13291248679161072, "learning_rate": 8.696137850060739e-05, "loss": 0.0368, "step": 43750 }, { "epoch": 0.8744654490228209, "grad_norm": 0.1196310818195343, "learning_rate": 8.695441809433405e-05, "loss": 0.0372, "step": 43760 }, { "epoch": 0.8746652811638224, "grad_norm": 0.11359766870737076, "learning_rate": 8.694745610943584e-05, "loss": 0.0358, "step": 43770 }, { "epoch": 0.874865113304824, "grad_norm": 0.12349526584148407, "learning_rate": 8.694049254621019e-05, "loss": 0.033, "step": 43780 }, { "epoch": 0.8750649454458255, "grad_norm": 0.1271304339170456, "learning_rate": 8.693352740495458e-05, "loss": 0.0298, "step": 43790 }, { "epoch": 0.8752647775868271, "grad_norm": 0.15112926065921783, "learning_rate": 8.69265606859665e-05, "loss": 0.038, "step": 43800 }, { "epoch": 0.8754646097278286, "grad_norm": 0.12390504032373428, "learning_rate": 8.691959238954359e-05, "loss": 0.0408, "step": 43810 }, { "epoch": 0.8756644418688302, "grad_norm": 0.12434707581996918, "learning_rate": 8.691262251598352e-05, "loss": 0.0371, "step": 43820 }, { "epoch": 0.8758642740098317, "grad_norm": 0.1501425951719284, "learning_rate": 8.690565106558404e-05, "loss": 0.0301, "step": 43830 }, { "epoch": 0.8760641061508333, "grad_norm": 0.16769500076770782, "learning_rate": 8.689867803864293e-05, "loss": 0.0311, "step": 43840 }, { "epoch": 0.8762639382918349, "grad_norm": 0.159663587808609, "learning_rate": 8.689170343545807e-05, "loss": 0.0467, "step": 43850 }, { "epoch": 0.8764637704328364, "grad_norm": 0.16098569333553314, "learning_rate": 8.688472725632743e-05, "loss": 0.0446, "step": 43860 }, { "epoch": 0.876663602573838, "grad_norm": 0.17428717017173767, "learning_rate": 8.687774950154899e-05, "loss": 0.0439, "step": 43870 }, { "epoch": 0.8768634347148395, "grad_norm": 0.15915410220623016, "learning_rate": 8.687077017142085e-05, "loss": 0.0415, "step": 43880 }, { "epoch": 0.8770632668558411, "grad_norm": 0.15082933008670807, "learning_rate": 8.686378926624112e-05, "loss": 0.0354, "step": 43890 }, { "epoch": 0.8772630989968426, "grad_norm": 0.14712287485599518, "learning_rate": 8.685680678630806e-05, "loss": 0.0359, "step": 43900 }, { "epoch": 0.8774629311378442, "grad_norm": 0.12532755732536316, "learning_rate": 8.68498227319199e-05, "loss": 0.0421, "step": 43910 }, { "epoch": 0.8776627632788457, "grad_norm": 0.11491920053958893, "learning_rate": 8.684283710337501e-05, "loss": 0.0279, "step": 43920 }, { "epoch": 0.8778625954198473, "grad_norm": 0.10926753282546997, "learning_rate": 8.683584990097182e-05, "loss": 0.0264, "step": 43930 }, { "epoch": 0.8780624275608488, "grad_norm": 0.18097807466983795, "learning_rate": 8.682886112500877e-05, "loss": 0.0338, "step": 43940 }, { "epoch": 0.8782622597018505, "grad_norm": 0.14174433052539825, "learning_rate": 8.682187077578445e-05, "loss": 0.0382, "step": 43950 }, { "epoch": 0.878462091842852, "grad_norm": 0.14950859546661377, "learning_rate": 8.681487885359745e-05, "loss": 0.0526, "step": 43960 }, { "epoch": 0.8786619239838536, "grad_norm": 0.1505972146987915, "learning_rate": 8.680788535874644e-05, "loss": 0.0395, "step": 43970 }, { "epoch": 0.8788617561248552, "grad_norm": 0.14670829474925995, "learning_rate": 8.68008902915302e-05, "loss": 0.0335, "step": 43980 }, { "epoch": 0.8790615882658567, "grad_norm": 0.1294131577014923, "learning_rate": 8.679389365224753e-05, "loss": 0.0335, "step": 43990 }, { "epoch": 0.8792614204068583, "grad_norm": 0.14119692146778107, "learning_rate": 8.678689544119731e-05, "loss": 0.0397, "step": 44000 }, { "epoch": 0.8794612525478598, "grad_norm": 0.18955788016319275, "learning_rate": 8.677989565867851e-05, "loss": 0.0384, "step": 44010 }, { "epoch": 0.8796610846888614, "grad_norm": 0.1322200894355774, "learning_rate": 8.677289430499012e-05, "loss": 0.0348, "step": 44020 }, { "epoch": 0.8798609168298629, "grad_norm": 0.14344006776809692, "learning_rate": 8.676589138043125e-05, "loss": 0.0411, "step": 44030 }, { "epoch": 0.8800607489708645, "grad_norm": 0.10993576049804688, "learning_rate": 8.675888688530102e-05, "loss": 0.0343, "step": 44040 }, { "epoch": 0.880260581111866, "grad_norm": 0.12793199717998505, "learning_rate": 8.675188081989869e-05, "loss": 0.0405, "step": 44050 }, { "epoch": 0.8804604132528676, "grad_norm": 0.1261787861585617, "learning_rate": 8.674487318452353e-05, "loss": 0.0387, "step": 44060 }, { "epoch": 0.8806602453938691, "grad_norm": 0.11983110755681992, "learning_rate": 8.673786397947488e-05, "loss": 0.0348, "step": 44070 }, { "epoch": 0.8808600775348707, "grad_norm": 0.1472233384847641, "learning_rate": 8.673085320505219e-05, "loss": 0.0382, "step": 44080 }, { "epoch": 0.8810599096758722, "grad_norm": 0.11886021494865417, "learning_rate": 8.67238408615549e-05, "loss": 0.0357, "step": 44090 }, { "epoch": 0.8812597418168738, "grad_norm": 0.15840992331504822, "learning_rate": 8.67168269492826e-05, "loss": 0.0386, "step": 44100 }, { "epoch": 0.8814595739578753, "grad_norm": 0.16378380358219147, "learning_rate": 8.67098114685349e-05, "loss": 0.0384, "step": 44110 }, { "epoch": 0.8816594060988769, "grad_norm": 0.12140829861164093, "learning_rate": 8.670279441961151e-05, "loss": 0.0259, "step": 44120 }, { "epoch": 0.8818592382398786, "grad_norm": 0.1345529854297638, "learning_rate": 8.669577580281217e-05, "loss": 0.0291, "step": 44130 }, { "epoch": 0.88205907038088, "grad_norm": 0.13737113773822784, "learning_rate": 8.668875561843668e-05, "loss": 0.0364, "step": 44140 }, { "epoch": 0.8822589025218817, "grad_norm": 0.12979602813720703, "learning_rate": 8.668173386678495e-05, "loss": 0.0446, "step": 44150 }, { "epoch": 0.8824587346628832, "grad_norm": 0.16838613152503967, "learning_rate": 8.667471054815695e-05, "loss": 0.035, "step": 44160 }, { "epoch": 0.8826585668038848, "grad_norm": 0.197299987077713, "learning_rate": 8.666768566285268e-05, "loss": 0.0416, "step": 44170 }, { "epoch": 0.8828583989448863, "grad_norm": 0.12213285267353058, "learning_rate": 8.666065921117224e-05, "loss": 0.0337, "step": 44180 }, { "epoch": 0.8830582310858879, "grad_norm": 0.13541238009929657, "learning_rate": 8.665363119341578e-05, "loss": 0.0331, "step": 44190 }, { "epoch": 0.8832580632268894, "grad_norm": 0.12752707302570343, "learning_rate": 8.664660160988354e-05, "loss": 0.0398, "step": 44200 }, { "epoch": 0.883457895367891, "grad_norm": 0.12873397767543793, "learning_rate": 8.66395704608758e-05, "loss": 0.0365, "step": 44210 }, { "epoch": 0.8836577275088925, "grad_norm": 0.14507277309894562, "learning_rate": 8.663253774669291e-05, "loss": 0.0462, "step": 44220 }, { "epoch": 0.8838575596498941, "grad_norm": 0.19146087765693665, "learning_rate": 8.662550346763532e-05, "loss": 0.0347, "step": 44230 }, { "epoch": 0.8840573917908956, "grad_norm": 0.1295718252658844, "learning_rate": 8.661846762400349e-05, "loss": 0.0319, "step": 44240 }, { "epoch": 0.8842572239318972, "grad_norm": 0.12752297520637512, "learning_rate": 8.661143021609799e-05, "loss": 0.0374, "step": 44250 }, { "epoch": 0.8844570560728988, "grad_norm": 0.15977072715759277, "learning_rate": 8.660439124421946e-05, "loss": 0.0285, "step": 44260 }, { "epoch": 0.8846568882139003, "grad_norm": 0.1209459975361824, "learning_rate": 8.659735070866859e-05, "loss": 0.0324, "step": 44270 }, { "epoch": 0.8848567203549019, "grad_norm": 0.15419752895832062, "learning_rate": 8.659030860974611e-05, "loss": 0.04, "step": 44280 }, { "epoch": 0.8850565524959034, "grad_norm": 0.12174860388040543, "learning_rate": 8.658326494775287e-05, "loss": 0.0329, "step": 44290 }, { "epoch": 0.885256384636905, "grad_norm": 0.15173357725143433, "learning_rate": 8.657621972298975e-05, "loss": 0.0422, "step": 44300 }, { "epoch": 0.8854562167779065, "grad_norm": 0.10965167731046677, "learning_rate": 8.656917293575773e-05, "loss": 0.0391, "step": 44310 }, { "epoch": 0.8856560489189081, "grad_norm": 0.14363163709640503, "learning_rate": 8.656212458635782e-05, "loss": 0.0423, "step": 44320 }, { "epoch": 0.8858558810599096, "grad_norm": 0.12464684247970581, "learning_rate": 8.655507467509112e-05, "loss": 0.0259, "step": 44330 }, { "epoch": 0.8860557132009113, "grad_norm": 0.11888401210308075, "learning_rate": 8.654802320225877e-05, "loss": 0.041, "step": 44340 }, { "epoch": 0.8862555453419128, "grad_norm": 0.1484326273202896, "learning_rate": 8.654097016816202e-05, "loss": 0.0346, "step": 44350 }, { "epoch": 0.8864553774829144, "grad_norm": 0.17216704785823822, "learning_rate": 8.653391557310214e-05, "loss": 0.0426, "step": 44360 }, { "epoch": 0.8866552096239159, "grad_norm": 0.12665113806724548, "learning_rate": 8.652685941738051e-05, "loss": 0.0422, "step": 44370 }, { "epoch": 0.8868550417649175, "grad_norm": 0.22227051854133606, "learning_rate": 8.651980170129856e-05, "loss": 0.0275, "step": 44380 }, { "epoch": 0.8870548739059191, "grad_norm": 0.13261951506137848, "learning_rate": 8.651274242515774e-05, "loss": 0.0356, "step": 44390 }, { "epoch": 0.8872547060469206, "grad_norm": 0.11523304134607315, "learning_rate": 8.650568158925967e-05, "loss": 0.0351, "step": 44400 }, { "epoch": 0.8874545381879222, "grad_norm": 0.16020503640174866, "learning_rate": 8.649861919390592e-05, "loss": 0.0274, "step": 44410 }, { "epoch": 0.8876543703289237, "grad_norm": 0.12354394048452377, "learning_rate": 8.649155523939822e-05, "loss": 0.034, "step": 44420 }, { "epoch": 0.8878542024699253, "grad_norm": 0.12816853821277618, "learning_rate": 8.648448972603832e-05, "loss": 0.0427, "step": 44430 }, { "epoch": 0.8880540346109268, "grad_norm": 0.13506948947906494, "learning_rate": 8.647742265412804e-05, "loss": 0.0431, "step": 44440 }, { "epoch": 0.8882538667519284, "grad_norm": 0.16265058517456055, "learning_rate": 8.647035402396928e-05, "loss": 0.0369, "step": 44450 }, { "epoch": 0.8884536988929299, "grad_norm": 0.17681922018527985, "learning_rate": 8.6463283835864e-05, "loss": 0.0357, "step": 44460 }, { "epoch": 0.8886535310339315, "grad_norm": 0.10134607553482056, "learning_rate": 8.64562120901142e-05, "loss": 0.0306, "step": 44470 }, { "epoch": 0.888853363174933, "grad_norm": 0.13267582654953003, "learning_rate": 8.6449138787022e-05, "loss": 0.041, "step": 44480 }, { "epoch": 0.8890531953159346, "grad_norm": 0.11019838601350784, "learning_rate": 8.644206392688956e-05, "loss": 0.0388, "step": 44490 }, { "epoch": 0.8892530274569361, "grad_norm": 0.13745497167110443, "learning_rate": 8.643498751001909e-05, "loss": 0.0369, "step": 44500 }, { "epoch": 0.8894528595979377, "grad_norm": 0.15373174846172333, "learning_rate": 8.642790953671286e-05, "loss": 0.0474, "step": 44510 }, { "epoch": 0.8896526917389392, "grad_norm": 0.12220500409603119, "learning_rate": 8.64208300072733e-05, "loss": 0.0341, "step": 44520 }, { "epoch": 0.8898525238799408, "grad_norm": 0.13954123854637146, "learning_rate": 8.641374892200276e-05, "loss": 0.0351, "step": 44530 }, { "epoch": 0.8900523560209425, "grad_norm": 0.17885243892669678, "learning_rate": 8.640666628120378e-05, "loss": 0.0348, "step": 44540 }, { "epoch": 0.890252188161944, "grad_norm": 0.12120195478200912, "learning_rate": 8.639958208517888e-05, "loss": 0.0415, "step": 44550 }, { "epoch": 0.8904520203029456, "grad_norm": 0.13180063664913177, "learning_rate": 8.63924963342307e-05, "loss": 0.0299, "step": 44560 }, { "epoch": 0.8906518524439471, "grad_norm": 0.16050755977630615, "learning_rate": 8.638540902866194e-05, "loss": 0.0331, "step": 44570 }, { "epoch": 0.8908516845849487, "grad_norm": 0.16483791172504425, "learning_rate": 8.637832016877535e-05, "loss": 0.0331, "step": 44580 }, { "epoch": 0.8910515167259502, "grad_norm": 0.14562706649303436, "learning_rate": 8.637122975487375e-05, "loss": 0.0337, "step": 44590 }, { "epoch": 0.8912513488669518, "grad_norm": 0.1282159388065338, "learning_rate": 8.636413778726002e-05, "loss": 0.0331, "step": 44600 }, { "epoch": 0.8914511810079533, "grad_norm": 0.15380404889583588, "learning_rate": 8.635704426623715e-05, "loss": 0.0351, "step": 44610 }, { "epoch": 0.8916510131489549, "grad_norm": 0.09046819806098938, "learning_rate": 8.634994919210813e-05, "loss": 0.0361, "step": 44620 }, { "epoch": 0.8918508452899564, "grad_norm": 0.15313033759593964, "learning_rate": 8.634285256517604e-05, "loss": 0.0375, "step": 44630 }, { "epoch": 0.892050677430958, "grad_norm": 0.1302160918712616, "learning_rate": 8.633575438574408e-05, "loss": 0.0413, "step": 44640 }, { "epoch": 0.8922505095719595, "grad_norm": 0.16365720331668854, "learning_rate": 8.632865465411541e-05, "loss": 0.0386, "step": 44650 }, { "epoch": 0.8924503417129611, "grad_norm": 0.17064961791038513, "learning_rate": 8.632155337059338e-05, "loss": 0.0402, "step": 44660 }, { "epoch": 0.8926501738539627, "grad_norm": 0.16976897418498993, "learning_rate": 8.63144505354813e-05, "loss": 0.0412, "step": 44670 }, { "epoch": 0.8928500059949642, "grad_norm": 0.14266498386859894, "learning_rate": 8.630734614908262e-05, "loss": 0.0364, "step": 44680 }, { "epoch": 0.8930498381359658, "grad_norm": 0.15530714392662048, "learning_rate": 8.63002402117008e-05, "loss": 0.0388, "step": 44690 }, { "epoch": 0.8932496702769673, "grad_norm": 0.15058228373527527, "learning_rate": 8.62931327236394e-05, "loss": 0.0328, "step": 44700 }, { "epoch": 0.8934495024179689, "grad_norm": 0.10804107785224915, "learning_rate": 8.628602368520204e-05, "loss": 0.0285, "step": 44710 }, { "epoch": 0.8936493345589704, "grad_norm": 0.11785779893398285, "learning_rate": 8.627891309669243e-05, "loss": 0.036, "step": 44720 }, { "epoch": 0.893849166699972, "grad_norm": 0.1280488818883896, "learning_rate": 8.627180095841429e-05, "loss": 0.0375, "step": 44730 }, { "epoch": 0.8940489988409736, "grad_norm": 0.09730537235736847, "learning_rate": 8.626468727067143e-05, "loss": 0.0439, "step": 44740 }, { "epoch": 0.8942488309819752, "grad_norm": 0.16697542369365692, "learning_rate": 8.625757203376776e-05, "loss": 0.0457, "step": 44750 }, { "epoch": 0.8944486631229767, "grad_norm": 0.14552119374275208, "learning_rate": 8.625045524800723e-05, "loss": 0.035, "step": 44760 }, { "epoch": 0.8946484952639783, "grad_norm": 0.15645888447761536, "learning_rate": 8.624333691369384e-05, "loss": 0.0355, "step": 44770 }, { "epoch": 0.8948483274049798, "grad_norm": 0.1373499482870102, "learning_rate": 8.623621703113167e-05, "loss": 0.0357, "step": 44780 }, { "epoch": 0.8950481595459814, "grad_norm": 0.113298200070858, "learning_rate": 8.622909560062487e-05, "loss": 0.04, "step": 44790 }, { "epoch": 0.8952479916869829, "grad_norm": 0.14465390145778656, "learning_rate": 8.622197262247767e-05, "loss": 0.0375, "step": 44800 }, { "epoch": 0.8954478238279845, "grad_norm": 0.12554597854614258, "learning_rate": 8.621484809699436e-05, "loss": 0.0388, "step": 44810 }, { "epoch": 0.8956476559689861, "grad_norm": 0.11281737685203552, "learning_rate": 8.620772202447924e-05, "loss": 0.0492, "step": 44820 }, { "epoch": 0.8958474881099876, "grad_norm": 0.12353368103504181, "learning_rate": 8.620059440523678e-05, "loss": 0.0351, "step": 44830 }, { "epoch": 0.8960473202509892, "grad_norm": 0.13365952670574188, "learning_rate": 8.61934652395714e-05, "loss": 0.0349, "step": 44840 }, { "epoch": 0.8962471523919907, "grad_norm": 0.2150227576494217, "learning_rate": 8.618633452778768e-05, "loss": 0.0361, "step": 44850 }, { "epoch": 0.8964469845329923, "grad_norm": 0.12433616816997528, "learning_rate": 8.617920227019024e-05, "loss": 0.0368, "step": 44860 }, { "epoch": 0.8966468166739938, "grad_norm": 0.11115478724241257, "learning_rate": 8.617206846708374e-05, "loss": 0.0321, "step": 44870 }, { "epoch": 0.8968466488149954, "grad_norm": 0.12659277021884918, "learning_rate": 8.616493311877292e-05, "loss": 0.0295, "step": 44880 }, { "epoch": 0.8970464809559969, "grad_norm": 0.1578478068113327, "learning_rate": 8.61577962255626e-05, "loss": 0.0456, "step": 44890 }, { "epoch": 0.8972463130969985, "grad_norm": 0.12027132511138916, "learning_rate": 8.615065778775763e-05, "loss": 0.0445, "step": 44900 }, { "epoch": 0.897446145238, "grad_norm": 0.14418786764144897, "learning_rate": 8.614351780566299e-05, "loss": 0.0404, "step": 44910 }, { "epoch": 0.8976459773790016, "grad_norm": 0.15884929895401, "learning_rate": 8.613637627958367e-05, "loss": 0.0422, "step": 44920 }, { "epoch": 0.8978458095200031, "grad_norm": 0.14364926517009735, "learning_rate": 8.612923320982472e-05, "loss": 0.0337, "step": 44930 }, { "epoch": 0.8980456416610048, "grad_norm": 0.13586099445819855, "learning_rate": 8.61220885966913e-05, "loss": 0.0365, "step": 44940 }, { "epoch": 0.8982454738020064, "grad_norm": 0.13958382606506348, "learning_rate": 8.611494244048863e-05, "loss": 0.0356, "step": 44950 }, { "epoch": 0.8984453059430079, "grad_norm": 0.1955508440732956, "learning_rate": 8.610779474152196e-05, "loss": 0.0402, "step": 44960 }, { "epoch": 0.8986451380840095, "grad_norm": 0.14900626242160797, "learning_rate": 8.610064550009663e-05, "loss": 0.0313, "step": 44970 }, { "epoch": 0.898844970225011, "grad_norm": 0.19876068830490112, "learning_rate": 8.609349471651804e-05, "loss": 0.0513, "step": 44980 }, { "epoch": 0.8990448023660126, "grad_norm": 0.10627710074186325, "learning_rate": 8.608634239109166e-05, "loss": 0.0332, "step": 44990 }, { "epoch": 0.8992446345070141, "grad_norm": 0.12795434892177582, "learning_rate": 8.607918852412304e-05, "loss": 0.0389, "step": 45000 }, { "epoch": 0.8994444666480157, "grad_norm": 0.11970420926809311, "learning_rate": 8.607203311591775e-05, "loss": 0.0296, "step": 45010 }, { "epoch": 0.8996442987890172, "grad_norm": 0.1253540962934494, "learning_rate": 8.606487616678147e-05, "loss": 0.036, "step": 45020 }, { "epoch": 0.8998441309300188, "grad_norm": 0.13912566006183624, "learning_rate": 8.605771767701995e-05, "loss": 0.037, "step": 45030 }, { "epoch": 0.9000439630710203, "grad_norm": 0.19020125269889832, "learning_rate": 8.605055764693895e-05, "loss": 0.0375, "step": 45040 }, { "epoch": 0.9002437952120219, "grad_norm": 0.16898703575134277, "learning_rate": 8.604339607684438e-05, "loss": 0.042, "step": 45050 }, { "epoch": 0.9004436273530234, "grad_norm": 0.12894195318222046, "learning_rate": 8.603623296704212e-05, "loss": 0.0368, "step": 45060 }, { "epoch": 0.900643459494025, "grad_norm": 0.13595344126224518, "learning_rate": 8.60290683178382e-05, "loss": 0.0306, "step": 45070 }, { "epoch": 0.9008432916350266, "grad_norm": 0.14553110301494598, "learning_rate": 8.602190212953867e-05, "loss": 0.0338, "step": 45080 }, { "epoch": 0.9010431237760281, "grad_norm": 0.11513495445251465, "learning_rate": 8.601473440244965e-05, "loss": 0.0351, "step": 45090 }, { "epoch": 0.9012429559170297, "grad_norm": 0.1276405155658722, "learning_rate": 8.600756513687734e-05, "loss": 0.029, "step": 45100 }, { "epoch": 0.9014427880580312, "grad_norm": 0.12027986347675323, "learning_rate": 8.600039433312799e-05, "loss": 0.033, "step": 45110 }, { "epoch": 0.9016426201990329, "grad_norm": 0.13255548477172852, "learning_rate": 8.599322199150793e-05, "loss": 0.0338, "step": 45120 }, { "epoch": 0.9018424523400344, "grad_norm": 0.13306760787963867, "learning_rate": 8.598604811232356e-05, "loss": 0.0346, "step": 45130 }, { "epoch": 0.902042284481036, "grad_norm": 0.12659171223640442, "learning_rate": 8.597887269588131e-05, "loss": 0.0428, "step": 45140 }, { "epoch": 0.9022421166220375, "grad_norm": 0.16412971913814545, "learning_rate": 8.597169574248772e-05, "loss": 0.028, "step": 45150 }, { "epoch": 0.9024419487630391, "grad_norm": 0.11760164797306061, "learning_rate": 8.596451725244936e-05, "loss": 0.0327, "step": 45160 }, { "epoch": 0.9026417809040406, "grad_norm": 0.1477205753326416, "learning_rate": 8.59573372260729e-05, "loss": 0.033, "step": 45170 }, { "epoch": 0.9028416130450422, "grad_norm": 0.12008636444807053, "learning_rate": 8.595015566366505e-05, "loss": 0.0315, "step": 45180 }, { "epoch": 0.9030414451860437, "grad_norm": 0.1507328748703003, "learning_rate": 8.59429725655326e-05, "loss": 0.028, "step": 45190 }, { "epoch": 0.9032412773270453, "grad_norm": 0.1309453248977661, "learning_rate": 8.593578793198238e-05, "loss": 0.0398, "step": 45200 }, { "epoch": 0.9034411094680468, "grad_norm": 0.15412116050720215, "learning_rate": 8.592860176332131e-05, "loss": 0.0345, "step": 45210 }, { "epoch": 0.9036409416090484, "grad_norm": 0.11711191385984421, "learning_rate": 8.592141405985638e-05, "loss": 0.0375, "step": 45220 }, { "epoch": 0.90384077375005, "grad_norm": 0.14200057089328766, "learning_rate": 8.591422482189465e-05, "loss": 0.0358, "step": 45230 }, { "epoch": 0.9040406058910515, "grad_norm": 0.13115233182907104, "learning_rate": 8.590703404974321e-05, "loss": 0.0328, "step": 45240 }, { "epoch": 0.9042404380320531, "grad_norm": 0.14679862558841705, "learning_rate": 8.589984174370922e-05, "loss": 0.0324, "step": 45250 }, { "epoch": 0.9044402701730546, "grad_norm": 0.13649925589561462, "learning_rate": 8.589264790409994e-05, "loss": 0.0281, "step": 45260 }, { "epoch": 0.9046401023140562, "grad_norm": 0.12171364575624466, "learning_rate": 8.588545253122268e-05, "loss": 0.0428, "step": 45270 }, { "epoch": 0.9048399344550577, "grad_norm": 0.1259157359600067, "learning_rate": 8.587825562538483e-05, "loss": 0.0349, "step": 45280 }, { "epoch": 0.9050397665960593, "grad_norm": 0.11488085240125656, "learning_rate": 8.58710571868938e-05, "loss": 0.0397, "step": 45290 }, { "epoch": 0.9052395987370608, "grad_norm": 0.13631482422351837, "learning_rate": 8.58638572160571e-05, "loss": 0.0324, "step": 45300 }, { "epoch": 0.9054394308780624, "grad_norm": 0.13379323482513428, "learning_rate": 8.58566557131823e-05, "loss": 0.0351, "step": 45310 }, { "epoch": 0.905639263019064, "grad_norm": 0.12165943533182144, "learning_rate": 8.584945267857705e-05, "loss": 0.0353, "step": 45320 }, { "epoch": 0.9058390951600656, "grad_norm": 0.13055747747421265, "learning_rate": 8.584224811254904e-05, "loss": 0.045, "step": 45330 }, { "epoch": 0.906038927301067, "grad_norm": 0.12432746589183807, "learning_rate": 8.583504201540601e-05, "loss": 0.0301, "step": 45340 }, { "epoch": 0.9062387594420687, "grad_norm": 0.1295502930879593, "learning_rate": 8.582783438745585e-05, "loss": 0.0332, "step": 45350 }, { "epoch": 0.9064385915830703, "grad_norm": 0.1057615578174591, "learning_rate": 8.58206252290064e-05, "loss": 0.0408, "step": 45360 }, { "epoch": 0.9066384237240718, "grad_norm": 0.20100750029087067, "learning_rate": 8.581341454036566e-05, "loss": 0.0385, "step": 45370 }, { "epoch": 0.9068382558650734, "grad_norm": 0.1509818285703659, "learning_rate": 8.580620232184164e-05, "loss": 0.0419, "step": 45380 }, { "epoch": 0.9070380880060749, "grad_norm": 0.1271679699420929, "learning_rate": 8.579898857374242e-05, "loss": 0.0335, "step": 45390 }, { "epoch": 0.9072379201470765, "grad_norm": 0.15519775450229645, "learning_rate": 8.579177329637617e-05, "loss": 0.0465, "step": 45400 }, { "epoch": 0.907437752288078, "grad_norm": 0.12913011014461517, "learning_rate": 8.578455649005113e-05, "loss": 0.0374, "step": 45410 }, { "epoch": 0.9076375844290796, "grad_norm": 0.16245026886463165, "learning_rate": 8.577733815507558e-05, "loss": 0.0357, "step": 45420 }, { "epoch": 0.9078374165700811, "grad_norm": 0.15488989651203156, "learning_rate": 8.577011829175787e-05, "loss": 0.0347, "step": 45430 }, { "epoch": 0.9080372487110827, "grad_norm": 0.1280316710472107, "learning_rate": 8.576289690040641e-05, "loss": 0.0411, "step": 45440 }, { "epoch": 0.9082370808520842, "grad_norm": 0.12938548624515533, "learning_rate": 8.575567398132968e-05, "loss": 0.0469, "step": 45450 }, { "epoch": 0.9084369129930858, "grad_norm": 0.13670986890792847, "learning_rate": 8.574844953483627e-05, "loss": 0.0346, "step": 45460 }, { "epoch": 0.9086367451340873, "grad_norm": 0.17591789364814758, "learning_rate": 8.574122356123476e-05, "loss": 0.0416, "step": 45470 }, { "epoch": 0.9088365772750889, "grad_norm": 0.1657690405845642, "learning_rate": 8.573399606083383e-05, "loss": 0.0328, "step": 45480 }, { "epoch": 0.9090364094160904, "grad_norm": 0.14812645316123962, "learning_rate": 8.572676703394223e-05, "loss": 0.0359, "step": 45490 }, { "epoch": 0.909236241557092, "grad_norm": 0.15129460394382477, "learning_rate": 8.571953648086879e-05, "loss": 0.04, "step": 45500 }, { "epoch": 0.9094360736980936, "grad_norm": 0.15697839856147766, "learning_rate": 8.571230440192235e-05, "loss": 0.0405, "step": 45510 }, { "epoch": 0.9096359058390951, "grad_norm": 0.13991901278495789, "learning_rate": 8.570507079741189e-05, "loss": 0.0355, "step": 45520 }, { "epoch": 0.9098357379800968, "grad_norm": 0.14287105202674866, "learning_rate": 8.56978356676464e-05, "loss": 0.0355, "step": 45530 }, { "epoch": 0.9100355701210983, "grad_norm": 0.15176641941070557, "learning_rate": 8.569059901293493e-05, "loss": 0.0394, "step": 45540 }, { "epoch": 0.9102354022620999, "grad_norm": 0.12891174852848053, "learning_rate": 8.568336083358664e-05, "loss": 0.0341, "step": 45550 }, { "epoch": 0.9104352344031014, "grad_norm": 0.1283358782529831, "learning_rate": 8.567612112991075e-05, "loss": 0.0434, "step": 45560 }, { "epoch": 0.910635066544103, "grad_norm": 0.12183962017297745, "learning_rate": 8.566887990221648e-05, "loss": 0.0472, "step": 45570 }, { "epoch": 0.9108348986851045, "grad_norm": 0.11533801257610321, "learning_rate": 8.56616371508132e-05, "loss": 0.032, "step": 45580 }, { "epoch": 0.9110347308261061, "grad_norm": 0.13648676872253418, "learning_rate": 8.565439287601029e-05, "loss": 0.0396, "step": 45590 }, { "epoch": 0.9112345629671076, "grad_norm": 0.15462175011634827, "learning_rate": 8.564714707811721e-05, "loss": 0.0393, "step": 45600 }, { "epoch": 0.9114343951081092, "grad_norm": 0.18647924065589905, "learning_rate": 8.563989975744349e-05, "loss": 0.0391, "step": 45610 }, { "epoch": 0.9116342272491107, "grad_norm": 0.13550475239753723, "learning_rate": 8.563265091429873e-05, "loss": 0.0374, "step": 45620 }, { "epoch": 0.9118340593901123, "grad_norm": 0.16886848211288452, "learning_rate": 8.56254005489926e-05, "loss": 0.0343, "step": 45630 }, { "epoch": 0.9120338915311139, "grad_norm": 0.11949636787176132, "learning_rate": 8.561814866183477e-05, "loss": 0.0362, "step": 45640 }, { "epoch": 0.9122337236721154, "grad_norm": 0.14341352880001068, "learning_rate": 8.56108952531351e-05, "loss": 0.0314, "step": 45650 }, { "epoch": 0.912433555813117, "grad_norm": 0.14479027688503265, "learning_rate": 8.560364032320338e-05, "loss": 0.0349, "step": 45660 }, { "epoch": 0.9126333879541185, "grad_norm": 0.13353215157985687, "learning_rate": 8.559638387234956e-05, "loss": 0.0351, "step": 45670 }, { "epoch": 0.9128332200951201, "grad_norm": 0.1339893639087677, "learning_rate": 8.558912590088359e-05, "loss": 0.0344, "step": 45680 }, { "epoch": 0.9130330522361216, "grad_norm": 0.1718415915966034, "learning_rate": 8.558186640911556e-05, "loss": 0.031, "step": 45690 }, { "epoch": 0.9132328843771232, "grad_norm": 0.11531104892492294, "learning_rate": 8.557460539735558e-05, "loss": 0.0356, "step": 45700 }, { "epoch": 0.9134327165181247, "grad_norm": 0.12305837124586105, "learning_rate": 8.556734286591378e-05, "loss": 0.0388, "step": 45710 }, { "epoch": 0.9136325486591264, "grad_norm": 0.137867271900177, "learning_rate": 8.556007881510044e-05, "loss": 0.0393, "step": 45720 }, { "epoch": 0.9138323808001279, "grad_norm": 0.10968240350484848, "learning_rate": 8.555281324522586e-05, "loss": 0.0307, "step": 45730 }, { "epoch": 0.9140322129411295, "grad_norm": 0.14111757278442383, "learning_rate": 8.554554615660042e-05, "loss": 0.0303, "step": 45740 }, { "epoch": 0.914232045082131, "grad_norm": 0.13628587126731873, "learning_rate": 8.553827754953455e-05, "loss": 0.0405, "step": 45750 }, { "epoch": 0.9144318772231326, "grad_norm": 0.13376906514167786, "learning_rate": 8.553100742433875e-05, "loss": 0.0338, "step": 45760 }, { "epoch": 0.9146317093641342, "grad_norm": 0.10501989722251892, "learning_rate": 8.552373578132356e-05, "loss": 0.038, "step": 45770 }, { "epoch": 0.9148315415051357, "grad_norm": 0.15994885563850403, "learning_rate": 8.551646262079966e-05, "loss": 0.0406, "step": 45780 }, { "epoch": 0.9150313736461373, "grad_norm": 0.10676916688680649, "learning_rate": 8.550918794307772e-05, "loss": 0.0302, "step": 45790 }, { "epoch": 0.9152312057871388, "grad_norm": 0.17443597316741943, "learning_rate": 8.550191174846849e-05, "loss": 0.0373, "step": 45800 }, { "epoch": 0.9154310379281404, "grad_norm": 0.11986896395683289, "learning_rate": 8.549463403728282e-05, "loss": 0.0364, "step": 45810 }, { "epoch": 0.9156308700691419, "grad_norm": 0.1783038079738617, "learning_rate": 8.548735480983158e-05, "loss": 0.0432, "step": 45820 }, { "epoch": 0.9158307022101435, "grad_norm": 0.15643306076526642, "learning_rate": 8.548007406642575e-05, "loss": 0.0396, "step": 45830 }, { "epoch": 0.916030534351145, "grad_norm": 0.18304060399532318, "learning_rate": 8.547279180737632e-05, "loss": 0.033, "step": 45840 }, { "epoch": 0.9162303664921466, "grad_norm": 0.20200371742248535, "learning_rate": 8.546550803299441e-05, "loss": 0.0415, "step": 45850 }, { "epoch": 0.9164301986331481, "grad_norm": 0.2542573809623718, "learning_rate": 8.545822274359114e-05, "loss": 0.0335, "step": 45860 }, { "epoch": 0.9166300307741497, "grad_norm": 0.17049582302570343, "learning_rate": 8.545093593947772e-05, "loss": 0.032, "step": 45870 }, { "epoch": 0.9168298629151512, "grad_norm": 0.1616310030221939, "learning_rate": 8.544364762096546e-05, "loss": 0.0316, "step": 45880 }, { "epoch": 0.9170296950561528, "grad_norm": 0.1417606920003891, "learning_rate": 8.543635778836567e-05, "loss": 0.0378, "step": 45890 }, { "epoch": 0.9172295271971543, "grad_norm": 0.15101474523544312, "learning_rate": 8.54290664419898e-05, "loss": 0.0379, "step": 45900 }, { "epoch": 0.917429359338156, "grad_norm": 0.1396830677986145, "learning_rate": 8.542177358214927e-05, "loss": 0.034, "step": 45910 }, { "epoch": 0.9176291914791576, "grad_norm": 0.14002612233161926, "learning_rate": 8.541447920915564e-05, "loss": 0.0361, "step": 45920 }, { "epoch": 0.9178290236201591, "grad_norm": 0.139933779835701, "learning_rate": 8.540718332332054e-05, "loss": 0.0314, "step": 45930 }, { "epoch": 0.9180288557611607, "grad_norm": 0.15337370336055756, "learning_rate": 8.53998859249556e-05, "loss": 0.0312, "step": 45940 }, { "epoch": 0.9182286879021622, "grad_norm": 0.13157489895820618, "learning_rate": 8.539258701437257e-05, "loss": 0.0405, "step": 45950 }, { "epoch": 0.9184285200431638, "grad_norm": 0.12144653499126434, "learning_rate": 8.538528659188324e-05, "loss": 0.0373, "step": 45960 }, { "epoch": 0.9186283521841653, "grad_norm": 0.17653828859329224, "learning_rate": 8.537798465779948e-05, "loss": 0.0322, "step": 45970 }, { "epoch": 0.9188281843251669, "grad_norm": 0.12572965025901794, "learning_rate": 8.537068121243319e-05, "loss": 0.0347, "step": 45980 }, { "epoch": 0.9190280164661684, "grad_norm": 0.13437668979167938, "learning_rate": 8.536337625609638e-05, "loss": 0.0391, "step": 45990 }, { "epoch": 0.91922784860717, "grad_norm": 0.16541093587875366, "learning_rate": 8.535606978910111e-05, "loss": 0.0383, "step": 46000 }, { "epoch": 0.9194276807481715, "grad_norm": 0.16809162497520447, "learning_rate": 8.534876181175948e-05, "loss": 0.037, "step": 46010 }, { "epoch": 0.9196275128891731, "grad_norm": 0.13938379287719727, "learning_rate": 8.534145232438369e-05, "loss": 0.0425, "step": 46020 }, { "epoch": 0.9198273450301746, "grad_norm": 0.2146816998720169, "learning_rate": 8.5334141327286e-05, "loss": 0.0378, "step": 46030 }, { "epoch": 0.9200271771711762, "grad_norm": 0.14504924416542053, "learning_rate": 8.532682882077867e-05, "loss": 0.0376, "step": 46040 }, { "epoch": 0.9202270093121778, "grad_norm": 0.1396707147359848, "learning_rate": 8.53195148051741e-05, "loss": 0.0384, "step": 46050 }, { "epoch": 0.9204268414531793, "grad_norm": 0.11360558867454529, "learning_rate": 8.531219928078477e-05, "loss": 0.0383, "step": 46060 }, { "epoch": 0.9206266735941809, "grad_norm": 0.1271330863237381, "learning_rate": 8.530488224792316e-05, "loss": 0.034, "step": 46070 }, { "epoch": 0.9208265057351824, "grad_norm": 0.14012396335601807, "learning_rate": 8.529756370690183e-05, "loss": 0.043, "step": 46080 }, { "epoch": 0.921026337876184, "grad_norm": 0.11736477166414261, "learning_rate": 8.529024365803342e-05, "loss": 0.0408, "step": 46090 }, { "epoch": 0.9212261700171855, "grad_norm": 0.1386353075504303, "learning_rate": 8.528292210163063e-05, "loss": 0.0483, "step": 46100 }, { "epoch": 0.9214260021581872, "grad_norm": 0.13402159512043, "learning_rate": 8.527559903800624e-05, "loss": 0.0293, "step": 46110 }, { "epoch": 0.9216258342991887, "grad_norm": 0.13764265179634094, "learning_rate": 8.526827446747303e-05, "loss": 0.0394, "step": 46120 }, { "epoch": 0.9218256664401903, "grad_norm": 0.15745986998081207, "learning_rate": 8.526094839034395e-05, "loss": 0.0347, "step": 46130 }, { "epoch": 0.9220254985811918, "grad_norm": 0.14357362687587738, "learning_rate": 8.525362080693194e-05, "loss": 0.0444, "step": 46140 }, { "epoch": 0.9222253307221934, "grad_norm": 0.1387181133031845, "learning_rate": 8.524629171754998e-05, "loss": 0.0367, "step": 46150 }, { "epoch": 0.9224251628631949, "grad_norm": 0.1317170411348343, "learning_rate": 8.523896112251122e-05, "loss": 0.0396, "step": 46160 }, { "epoch": 0.9226249950041965, "grad_norm": 0.133908212184906, "learning_rate": 8.523162902212876e-05, "loss": 0.033, "step": 46170 }, { "epoch": 0.922824827145198, "grad_norm": 0.1689387708902359, "learning_rate": 8.522429541671583e-05, "loss": 0.0391, "step": 46180 }, { "epoch": 0.9230246592861996, "grad_norm": 0.14306092262268066, "learning_rate": 8.521696030658571e-05, "loss": 0.0331, "step": 46190 }, { "epoch": 0.9232244914272012, "grad_norm": 0.14153288304805756, "learning_rate": 8.520962369205174e-05, "loss": 0.0524, "step": 46200 }, { "epoch": 0.9234243235682027, "grad_norm": 0.14640529453754425, "learning_rate": 8.520228557342733e-05, "loss": 0.0389, "step": 46210 }, { "epoch": 0.9236241557092043, "grad_norm": 0.14367038011550903, "learning_rate": 8.519494595102595e-05, "loss": 0.0365, "step": 46220 }, { "epoch": 0.9238239878502058, "grad_norm": 0.1382075995206833, "learning_rate": 8.518760482516113e-05, "loss": 0.0402, "step": 46230 }, { "epoch": 0.9240238199912074, "grad_norm": 0.15413925051689148, "learning_rate": 8.518026219614648e-05, "loss": 0.0354, "step": 46240 }, { "epoch": 0.9242236521322089, "grad_norm": 0.10265125334262848, "learning_rate": 8.517291806429566e-05, "loss": 0.0347, "step": 46250 }, { "epoch": 0.9244234842732105, "grad_norm": 0.14256052672863007, "learning_rate": 8.516557242992238e-05, "loss": 0.0368, "step": 46260 }, { "epoch": 0.924623316414212, "grad_norm": 0.16054360568523407, "learning_rate": 8.515822529334046e-05, "loss": 0.0327, "step": 46270 }, { "epoch": 0.9248231485552136, "grad_norm": 0.1249866858124733, "learning_rate": 8.515087665486374e-05, "loss": 0.0371, "step": 46280 }, { "epoch": 0.9250229806962151, "grad_norm": 0.11836284399032593, "learning_rate": 8.514352651480615e-05, "loss": 0.0364, "step": 46290 }, { "epoch": 0.9252228128372167, "grad_norm": 0.10111790150403976, "learning_rate": 8.513617487348165e-05, "loss": 0.0318, "step": 46300 }, { "epoch": 0.9254226449782182, "grad_norm": 0.11184733361005783, "learning_rate": 8.512882173120434e-05, "loss": 0.0446, "step": 46310 }, { "epoch": 0.9256224771192199, "grad_norm": 0.13426876068115234, "learning_rate": 8.512146708828825e-05, "loss": 0.0392, "step": 46320 }, { "epoch": 0.9258223092602215, "grad_norm": 0.17554743587970734, "learning_rate": 8.511411094504765e-05, "loss": 0.0393, "step": 46330 }, { "epoch": 0.926022141401223, "grad_norm": 0.11337932199239731, "learning_rate": 8.510675330179672e-05, "loss": 0.035, "step": 46340 }, { "epoch": 0.9262219735422246, "grad_norm": 0.12790462374687195, "learning_rate": 8.509939415884977e-05, "loss": 0.035, "step": 46350 }, { "epoch": 0.9264218056832261, "grad_norm": 0.1177750900387764, "learning_rate": 8.509203351652121e-05, "loss": 0.0393, "step": 46360 }, { "epoch": 0.9266216378242277, "grad_norm": 0.14099673926830292, "learning_rate": 8.508467137512542e-05, "loss": 0.0351, "step": 46370 }, { "epoch": 0.9268214699652292, "grad_norm": 0.09324751049280167, "learning_rate": 8.507730773497693e-05, "loss": 0.0254, "step": 46380 }, { "epoch": 0.9270213021062308, "grad_norm": 0.12315002083778381, "learning_rate": 8.50699425963903e-05, "loss": 0.0439, "step": 46390 }, { "epoch": 0.9272211342472323, "grad_norm": 0.13912738859653473, "learning_rate": 8.506257595968013e-05, "loss": 0.035, "step": 46400 }, { "epoch": 0.9274209663882339, "grad_norm": 0.15355156362056732, "learning_rate": 8.505520782516114e-05, "loss": 0.0463, "step": 46410 }, { "epoch": 0.9276207985292354, "grad_norm": 0.15637746453285217, "learning_rate": 8.504783819314807e-05, "loss": 0.0428, "step": 46420 }, { "epoch": 0.927820630670237, "grad_norm": 0.1607939600944519, "learning_rate": 8.504046706395573e-05, "loss": 0.0371, "step": 46430 }, { "epoch": 0.9280204628112385, "grad_norm": 0.1428728848695755, "learning_rate": 8.503309443789901e-05, "loss": 0.0333, "step": 46440 }, { "epoch": 0.9282202949522401, "grad_norm": 0.12307550013065338, "learning_rate": 8.502572031529285e-05, "loss": 0.0494, "step": 46450 }, { "epoch": 0.9284201270932417, "grad_norm": 0.14423784613609314, "learning_rate": 8.501834469645228e-05, "loss": 0.0395, "step": 46460 }, { "epoch": 0.9286199592342432, "grad_norm": 0.17100904881954193, "learning_rate": 8.501096758169234e-05, "loss": 0.0346, "step": 46470 }, { "epoch": 0.9288197913752448, "grad_norm": 0.15870317816734314, "learning_rate": 8.500358897132818e-05, "loss": 0.0363, "step": 46480 }, { "epoch": 0.9290196235162463, "grad_norm": 0.14855439960956573, "learning_rate": 8.499620886567502e-05, "loss": 0.0395, "step": 46490 }, { "epoch": 0.929219455657248, "grad_norm": 0.14253295958042145, "learning_rate": 8.498882726504808e-05, "loss": 0.0467, "step": 46500 }, { "epoch": 0.9294192877982494, "grad_norm": 0.12863706052303314, "learning_rate": 8.498144416976274e-05, "loss": 0.0306, "step": 46510 }, { "epoch": 0.9296191199392511, "grad_norm": 0.19887182116508484, "learning_rate": 8.497405958013435e-05, "loss": 0.0538, "step": 46520 }, { "epoch": 0.9298189520802526, "grad_norm": 0.1554209589958191, "learning_rate": 8.496667349647841e-05, "loss": 0.0355, "step": 46530 }, { "epoch": 0.9300187842212542, "grad_norm": 0.13791094720363617, "learning_rate": 8.495928591911039e-05, "loss": 0.0362, "step": 46540 }, { "epoch": 0.9302186163622557, "grad_norm": 0.11384614557027817, "learning_rate": 8.495189684834592e-05, "loss": 0.0356, "step": 46550 }, { "epoch": 0.9304184485032573, "grad_norm": 0.12524667382240295, "learning_rate": 8.494450628450062e-05, "loss": 0.0376, "step": 46560 }, { "epoch": 0.9306182806442588, "grad_norm": 0.16018597781658173, "learning_rate": 8.493711422789018e-05, "loss": 0.0398, "step": 46570 }, { "epoch": 0.9308181127852604, "grad_norm": 0.18732155859470367, "learning_rate": 8.492972067883043e-05, "loss": 0.0373, "step": 46580 }, { "epoch": 0.9310179449262619, "grad_norm": 0.1319931447505951, "learning_rate": 8.492232563763717e-05, "loss": 0.0345, "step": 46590 }, { "epoch": 0.9312177770672635, "grad_norm": 0.12047378718852997, "learning_rate": 8.49149291046263e-05, "loss": 0.0364, "step": 46600 }, { "epoch": 0.9314176092082651, "grad_norm": 0.12749651074409485, "learning_rate": 8.490753108011382e-05, "loss": 0.0339, "step": 46610 }, { "epoch": 0.9316174413492666, "grad_norm": 0.1271173506975174, "learning_rate": 8.490013156441574e-05, "loss": 0.0366, "step": 46620 }, { "epoch": 0.9318172734902682, "grad_norm": 0.13241344690322876, "learning_rate": 8.489273055784814e-05, "loss": 0.0413, "step": 46630 }, { "epoch": 0.9320171056312697, "grad_norm": 0.11914807558059692, "learning_rate": 8.48853280607272e-05, "loss": 0.0286, "step": 46640 }, { "epoch": 0.9322169377722713, "grad_norm": 0.11817967146635056, "learning_rate": 8.487792407336913e-05, "loss": 0.0422, "step": 46650 }, { "epoch": 0.9324167699132728, "grad_norm": 0.1585497111082077, "learning_rate": 8.48705185960902e-05, "loss": 0.0309, "step": 46660 }, { "epoch": 0.9326166020542744, "grad_norm": 0.11900589615106583, "learning_rate": 8.486311162920679e-05, "loss": 0.0331, "step": 46670 }, { "epoch": 0.9328164341952759, "grad_norm": 0.19653140008449554, "learning_rate": 8.48557031730353e-05, "loss": 0.0408, "step": 46680 }, { "epoch": 0.9330162663362775, "grad_norm": 0.13053762912750244, "learning_rate": 8.48482932278922e-05, "loss": 0.0314, "step": 46690 }, { "epoch": 0.933216098477279, "grad_norm": 0.13191990554332733, "learning_rate": 8.484088179409404e-05, "loss": 0.0424, "step": 46700 }, { "epoch": 0.9334159306182807, "grad_norm": 0.11649293452501297, "learning_rate": 8.48334688719574e-05, "loss": 0.0278, "step": 46710 }, { "epoch": 0.9336157627592822, "grad_norm": 0.15941205620765686, "learning_rate": 8.482605446179897e-05, "loss": 0.0271, "step": 46720 }, { "epoch": 0.9338155949002838, "grad_norm": 0.1146465539932251, "learning_rate": 8.481863856393545e-05, "loss": 0.034, "step": 46730 }, { "epoch": 0.9340154270412854, "grad_norm": 0.14209496974945068, "learning_rate": 8.481122117868367e-05, "loss": 0.0358, "step": 46740 }, { "epoch": 0.9342152591822869, "grad_norm": 0.18496547639369965, "learning_rate": 8.480380230636048e-05, "loss": 0.0382, "step": 46750 }, { "epoch": 0.9344150913232885, "grad_norm": 0.12428922951221466, "learning_rate": 8.47963819472828e-05, "loss": 0.037, "step": 46760 }, { "epoch": 0.93461492346429, "grad_norm": 0.17103154957294464, "learning_rate": 8.478896010176758e-05, "loss": 0.038, "step": 46770 }, { "epoch": 0.9348147556052916, "grad_norm": 0.11338745802640915, "learning_rate": 8.47815367701319e-05, "loss": 0.0384, "step": 46780 }, { "epoch": 0.9350145877462931, "grad_norm": 0.14584384858608246, "learning_rate": 8.477411195269287e-05, "loss": 0.0408, "step": 46790 }, { "epoch": 0.9352144198872947, "grad_norm": 0.13317839801311493, "learning_rate": 8.476668564976767e-05, "loss": 0.0345, "step": 46800 }, { "epoch": 0.9354142520282962, "grad_norm": 0.13006262481212616, "learning_rate": 8.475925786167351e-05, "loss": 0.0381, "step": 46810 }, { "epoch": 0.9356140841692978, "grad_norm": 0.10918816924095154, "learning_rate": 8.47518285887277e-05, "loss": 0.0303, "step": 46820 }, { "epoch": 0.9358139163102993, "grad_norm": 0.1616542786359787, "learning_rate": 8.474439783124764e-05, "loss": 0.0404, "step": 46830 }, { "epoch": 0.9360137484513009, "grad_norm": 0.12548354268074036, "learning_rate": 8.473696558955072e-05, "loss": 0.0319, "step": 46840 }, { "epoch": 0.9362135805923024, "grad_norm": 0.13851775228977203, "learning_rate": 8.472953186395444e-05, "loss": 0.0395, "step": 46850 }, { "epoch": 0.936413412733304, "grad_norm": 0.10343668609857559, "learning_rate": 8.472209665477638e-05, "loss": 0.0283, "step": 46860 }, { "epoch": 0.9366132448743055, "grad_norm": 0.15798042714595795, "learning_rate": 8.47146599623341e-05, "loss": 0.0363, "step": 46870 }, { "epoch": 0.9368130770153071, "grad_norm": 0.13410131633281708, "learning_rate": 8.470722178694535e-05, "loss": 0.0351, "step": 46880 }, { "epoch": 0.9370129091563087, "grad_norm": 0.1323690414428711, "learning_rate": 8.469978212892784e-05, "loss": 0.0372, "step": 46890 }, { "epoch": 0.9372127412973102, "grad_norm": 0.12511086463928223, "learning_rate": 8.469234098859937e-05, "loss": 0.0336, "step": 46900 }, { "epoch": 0.9374125734383119, "grad_norm": 0.11032474786043167, "learning_rate": 8.468489836627784e-05, "loss": 0.0389, "step": 46910 }, { "epoch": 0.9376124055793134, "grad_norm": 0.18394148349761963, "learning_rate": 8.467745426228115e-05, "loss": 0.0347, "step": 46920 }, { "epoch": 0.937812237720315, "grad_norm": 0.14092254638671875, "learning_rate": 8.467000867692734e-05, "loss": 0.0323, "step": 46930 }, { "epoch": 0.9380120698613165, "grad_norm": 0.11991578340530396, "learning_rate": 8.466256161053444e-05, "loss": 0.0422, "step": 46940 }, { "epoch": 0.9382119020023181, "grad_norm": 0.12828317284584045, "learning_rate": 8.465511306342059e-05, "loss": 0.0385, "step": 46950 }, { "epoch": 0.9384117341433196, "grad_norm": 0.14680266380310059, "learning_rate": 8.464766303590396e-05, "loss": 0.0409, "step": 46960 }, { "epoch": 0.9386115662843212, "grad_norm": 0.14394551515579224, "learning_rate": 8.464021152830282e-05, "loss": 0.036, "step": 46970 }, { "epoch": 0.9388113984253227, "grad_norm": 0.11715132743120193, "learning_rate": 8.46327585409355e-05, "loss": 0.0407, "step": 46980 }, { "epoch": 0.9390112305663243, "grad_norm": 0.12587498128414154, "learning_rate": 8.462530407412033e-05, "loss": 0.0389, "step": 46990 }, { "epoch": 0.9392110627073258, "grad_norm": 0.10565998405218124, "learning_rate": 8.46178481281758e-05, "loss": 0.0381, "step": 47000 }, { "epoch": 0.9394108948483274, "grad_norm": 0.1484430879354477, "learning_rate": 8.461039070342039e-05, "loss": 0.0345, "step": 47010 }, { "epoch": 0.939610726989329, "grad_norm": 0.15270528197288513, "learning_rate": 8.460293180017265e-05, "loss": 0.0325, "step": 47020 }, { "epoch": 0.9398105591303305, "grad_norm": 0.12577149271965027, "learning_rate": 8.459547141875126e-05, "loss": 0.0417, "step": 47030 }, { "epoch": 0.9400103912713321, "grad_norm": 0.12085874378681183, "learning_rate": 8.458800955947486e-05, "loss": 0.0407, "step": 47040 }, { "epoch": 0.9402102234123336, "grad_norm": 0.1379084289073944, "learning_rate": 8.458054622266225e-05, "loss": 0.0341, "step": 47050 }, { "epoch": 0.9404100555533352, "grad_norm": 0.17731818556785583, "learning_rate": 8.457308140863221e-05, "loss": 0.033, "step": 47060 }, { "epoch": 0.9406098876943367, "grad_norm": 0.16040173172950745, "learning_rate": 8.456561511770367e-05, "loss": 0.0398, "step": 47070 }, { "epoch": 0.9408097198353383, "grad_norm": 0.1513967514038086, "learning_rate": 8.455814735019553e-05, "loss": 0.0388, "step": 47080 }, { "epoch": 0.9410095519763398, "grad_norm": 0.14464432001113892, "learning_rate": 8.455067810642683e-05, "loss": 0.0294, "step": 47090 }, { "epoch": 0.9412093841173415, "grad_norm": 0.16837236285209656, "learning_rate": 8.454320738671664e-05, "loss": 0.0293, "step": 47100 }, { "epoch": 0.941409216258343, "grad_norm": 0.14410488307476044, "learning_rate": 8.453573519138409e-05, "loss": 0.0344, "step": 47110 }, { "epoch": 0.9416090483993446, "grad_norm": 0.10015382617712021, "learning_rate": 8.452826152074836e-05, "loss": 0.0315, "step": 47120 }, { "epoch": 0.9418088805403461, "grad_norm": 0.12220777571201324, "learning_rate": 8.452078637512873e-05, "loss": 0.0375, "step": 47130 }, { "epoch": 0.9420087126813477, "grad_norm": 0.13685373961925507, "learning_rate": 8.451330975484455e-05, "loss": 0.033, "step": 47140 }, { "epoch": 0.9422085448223493, "grad_norm": 0.16296741366386414, "learning_rate": 8.450583166021515e-05, "loss": 0.0344, "step": 47150 }, { "epoch": 0.9424083769633508, "grad_norm": 0.13271179795265198, "learning_rate": 8.449835209156003e-05, "loss": 0.0359, "step": 47160 }, { "epoch": 0.9426082091043524, "grad_norm": 0.16018013656139374, "learning_rate": 8.449087104919866e-05, "loss": 0.0334, "step": 47170 }, { "epoch": 0.9428080412453539, "grad_norm": 0.21055421233177185, "learning_rate": 8.448338853345066e-05, "loss": 0.0459, "step": 47180 }, { "epoch": 0.9430078733863555, "grad_norm": 0.1715453714132309, "learning_rate": 8.447590454463565e-05, "loss": 0.0431, "step": 47190 }, { "epoch": 0.943207705527357, "grad_norm": 0.14739519357681274, "learning_rate": 8.446841908307333e-05, "loss": 0.037, "step": 47200 }, { "epoch": 0.9434075376683586, "grad_norm": 0.11703954637050629, "learning_rate": 8.446093214908345e-05, "loss": 0.0314, "step": 47210 }, { "epoch": 0.9436073698093601, "grad_norm": 0.16475366055965424, "learning_rate": 8.445344374298588e-05, "loss": 0.0466, "step": 47220 }, { "epoch": 0.9438072019503617, "grad_norm": 0.16344863176345825, "learning_rate": 8.444595386510048e-05, "loss": 0.0348, "step": 47230 }, { "epoch": 0.9440070340913632, "grad_norm": 0.16070273518562317, "learning_rate": 8.443846251574721e-05, "loss": 0.0383, "step": 47240 }, { "epoch": 0.9442068662323648, "grad_norm": 0.13804765045642853, "learning_rate": 8.443096969524607e-05, "loss": 0.0377, "step": 47250 }, { "epoch": 0.9444066983733663, "grad_norm": 0.13811346888542175, "learning_rate": 8.442347540391718e-05, "loss": 0.0312, "step": 47260 }, { "epoch": 0.9446065305143679, "grad_norm": 0.13715218007564545, "learning_rate": 8.441597964208063e-05, "loss": 0.0367, "step": 47270 }, { "epoch": 0.9448063626553694, "grad_norm": 0.15259288251399994, "learning_rate": 8.44084824100567e-05, "loss": 0.0496, "step": 47280 }, { "epoch": 0.945006194796371, "grad_norm": 0.12892845273017883, "learning_rate": 8.440098370816556e-05, "loss": 0.0314, "step": 47290 }, { "epoch": 0.9452060269373727, "grad_norm": 0.1699814349412918, "learning_rate": 8.439348353672763e-05, "loss": 0.0433, "step": 47300 }, { "epoch": 0.9454058590783742, "grad_norm": 0.14251290261745453, "learning_rate": 8.438598189606324e-05, "loss": 0.0396, "step": 47310 }, { "epoch": 0.9456056912193758, "grad_norm": 0.11658438295125961, "learning_rate": 8.43784787864929e-05, "loss": 0.0327, "step": 47320 }, { "epoch": 0.9458055233603773, "grad_norm": 0.1201145350933075, "learning_rate": 8.437097420833707e-05, "loss": 0.037, "step": 47330 }, { "epoch": 0.9460053555013789, "grad_norm": 0.15125182271003723, "learning_rate": 8.436346816191638e-05, "loss": 0.0439, "step": 47340 }, { "epoch": 0.9462051876423804, "grad_norm": 0.13622267544269562, "learning_rate": 8.435596064755147e-05, "loss": 0.0327, "step": 47350 }, { "epoch": 0.946405019783382, "grad_norm": 0.1628745198249817, "learning_rate": 8.434845166556302e-05, "loss": 0.0432, "step": 47360 }, { "epoch": 0.9466048519243835, "grad_norm": 0.13375422358512878, "learning_rate": 8.43409412162718e-05, "loss": 0.037, "step": 47370 }, { "epoch": 0.9468046840653851, "grad_norm": 0.12270785868167877, "learning_rate": 8.433342929999869e-05, "loss": 0.0376, "step": 47380 }, { "epoch": 0.9470045162063866, "grad_norm": 0.11141551285982132, "learning_rate": 8.432591591706454e-05, "loss": 0.0387, "step": 47390 }, { "epoch": 0.9472043483473882, "grad_norm": 0.12938527762889862, "learning_rate": 8.431840106779032e-05, "loss": 0.0327, "step": 47400 }, { "epoch": 0.9474041804883897, "grad_norm": 0.1326730102300644, "learning_rate": 8.431088475249705e-05, "loss": 0.0316, "step": 47410 }, { "epoch": 0.9476040126293913, "grad_norm": 0.1218651533126831, "learning_rate": 8.430336697150582e-05, "loss": 0.033, "step": 47420 }, { "epoch": 0.9478038447703929, "grad_norm": 0.14639267325401306, "learning_rate": 8.429584772513777e-05, "loss": 0.0381, "step": 47430 }, { "epoch": 0.9480036769113944, "grad_norm": 0.14133639633655548, "learning_rate": 8.42883270137141e-05, "loss": 0.0379, "step": 47440 }, { "epoch": 0.948203509052396, "grad_norm": 0.137239009141922, "learning_rate": 8.42808048375561e-05, "loss": 0.0428, "step": 47450 }, { "epoch": 0.9484033411933975, "grad_norm": 0.1220303401350975, "learning_rate": 8.427328119698507e-05, "loss": 0.0345, "step": 47460 }, { "epoch": 0.9486031733343991, "grad_norm": 0.13540661334991455, "learning_rate": 8.426575609232245e-05, "loss": 0.0351, "step": 47470 }, { "epoch": 0.9488030054754006, "grad_norm": 0.1063639372587204, "learning_rate": 8.425822952388967e-05, "loss": 0.0339, "step": 47480 }, { "epoch": 0.9490028376164023, "grad_norm": 0.14642493426799774, "learning_rate": 8.425070149200827e-05, "loss": 0.0304, "step": 47490 }, { "epoch": 0.9492026697574037, "grad_norm": 0.12502145767211914, "learning_rate": 8.424317199699981e-05, "loss": 0.046, "step": 47500 }, { "epoch": 0.9494025018984054, "grad_norm": 0.16903206706047058, "learning_rate": 8.423564103918595e-05, "loss": 0.0335, "step": 47510 }, { "epoch": 0.9496023340394069, "grad_norm": 0.14832113683223724, "learning_rate": 8.422810861888841e-05, "loss": 0.0443, "step": 47520 }, { "epoch": 0.9498021661804085, "grad_norm": 0.168512761592865, "learning_rate": 8.422057473642893e-05, "loss": 0.0396, "step": 47530 }, { "epoch": 0.95000199832141, "grad_norm": 0.14271801710128784, "learning_rate": 8.421303939212938e-05, "loss": 0.0297, "step": 47540 }, { "epoch": 0.9502018304624116, "grad_norm": 0.10620324313640594, "learning_rate": 8.420550258631164e-05, "loss": 0.0332, "step": 47550 }, { "epoch": 0.9504016626034131, "grad_norm": 0.13512714207172394, "learning_rate": 8.419796431929765e-05, "loss": 0.0409, "step": 47560 }, { "epoch": 0.9506014947444147, "grad_norm": 0.14552603662014008, "learning_rate": 8.419042459140947e-05, "loss": 0.0397, "step": 47570 }, { "epoch": 0.9508013268854163, "grad_norm": 0.15437065064907074, "learning_rate": 8.418288340296913e-05, "loss": 0.0411, "step": 47580 }, { "epoch": 0.9510011590264178, "grad_norm": 0.15454037487506866, "learning_rate": 8.417534075429884e-05, "loss": 0.0382, "step": 47590 }, { "epoch": 0.9512009911674194, "grad_norm": 0.10871664434671402, "learning_rate": 8.416779664572077e-05, "loss": 0.0327, "step": 47600 }, { "epoch": 0.9514008233084209, "grad_norm": 0.1338668018579483, "learning_rate": 8.416025107755719e-05, "loss": 0.0328, "step": 47610 }, { "epoch": 0.9516006554494225, "grad_norm": 0.17397522926330566, "learning_rate": 8.415270405013041e-05, "loss": 0.0453, "step": 47620 }, { "epoch": 0.951800487590424, "grad_norm": 0.13489040732383728, "learning_rate": 8.414515556376288e-05, "loss": 0.0354, "step": 47630 }, { "epoch": 0.9520003197314256, "grad_norm": 0.11939950287342072, "learning_rate": 8.413760561877704e-05, "loss": 0.0329, "step": 47640 }, { "epoch": 0.9522001518724271, "grad_norm": 0.17734456062316895, "learning_rate": 8.413005421549538e-05, "loss": 0.0416, "step": 47650 }, { "epoch": 0.9523999840134287, "grad_norm": 0.1451655924320221, "learning_rate": 8.412250135424052e-05, "loss": 0.0417, "step": 47660 }, { "epoch": 0.9525998161544302, "grad_norm": 0.10661252588033676, "learning_rate": 8.411494703533508e-05, "loss": 0.0343, "step": 47670 }, { "epoch": 0.9527996482954318, "grad_norm": 0.12104623019695282, "learning_rate": 8.410739125910178e-05, "loss": 0.034, "step": 47680 }, { "epoch": 0.9529994804364333, "grad_norm": 0.12185264378786087, "learning_rate": 8.409983402586337e-05, "loss": 0.048, "step": 47690 }, { "epoch": 0.953199312577435, "grad_norm": 0.2215588092803955, "learning_rate": 8.40922753359427e-05, "loss": 0.0402, "step": 47700 }, { "epoch": 0.9533991447184366, "grad_norm": 0.1459430605173111, "learning_rate": 8.408471518966265e-05, "loss": 0.0417, "step": 47710 }, { "epoch": 0.9535989768594381, "grad_norm": 0.18144582211971283, "learning_rate": 8.40771535873462e-05, "loss": 0.0414, "step": 47720 }, { "epoch": 0.9537988090004397, "grad_norm": 0.1609867513179779, "learning_rate": 8.406959052931634e-05, "loss": 0.0382, "step": 47730 }, { "epoch": 0.9539986411414412, "grad_norm": 0.1392185240983963, "learning_rate": 8.406202601589615e-05, "loss": 0.0402, "step": 47740 }, { "epoch": 0.9541984732824428, "grad_norm": 0.16154009103775024, "learning_rate": 8.40544600474088e-05, "loss": 0.0461, "step": 47750 }, { "epoch": 0.9543983054234443, "grad_norm": 0.1356787234544754, "learning_rate": 8.404689262417747e-05, "loss": 0.047, "step": 47760 }, { "epoch": 0.9545981375644459, "grad_norm": 0.168524369597435, "learning_rate": 8.403932374652544e-05, "loss": 0.0366, "step": 47770 }, { "epoch": 0.9547979697054474, "grad_norm": 0.1655426174402237, "learning_rate": 8.403175341477601e-05, "loss": 0.0354, "step": 47780 }, { "epoch": 0.954997801846449, "grad_norm": 0.15135924518108368, "learning_rate": 8.402418162925262e-05, "loss": 0.0345, "step": 47790 }, { "epoch": 0.9551976339874505, "grad_norm": 0.11850038170814514, "learning_rate": 8.401660839027867e-05, "loss": 0.0321, "step": 47800 }, { "epoch": 0.9553974661284521, "grad_norm": 0.1519705206155777, "learning_rate": 8.400903369817771e-05, "loss": 0.0358, "step": 47810 }, { "epoch": 0.9555972982694536, "grad_norm": 0.14819954335689545, "learning_rate": 8.400145755327332e-05, "loss": 0.0417, "step": 47820 }, { "epoch": 0.9557971304104552, "grad_norm": 0.13357940316200256, "learning_rate": 8.399387995588913e-05, "loss": 0.0357, "step": 47830 }, { "epoch": 0.9559969625514568, "grad_norm": 0.12277667969465256, "learning_rate": 8.398630090634881e-05, "loss": 0.0471, "step": 47840 }, { "epoch": 0.9561967946924583, "grad_norm": 0.16105835139751434, "learning_rate": 8.397872040497617e-05, "loss": 0.0375, "step": 47850 }, { "epoch": 0.9563966268334599, "grad_norm": 0.17497752606868744, "learning_rate": 8.397113845209502e-05, "loss": 0.0403, "step": 47860 }, { "epoch": 0.9565964589744614, "grad_norm": 0.15607960522174835, "learning_rate": 8.396355504802922e-05, "loss": 0.0358, "step": 47870 }, { "epoch": 0.956796291115463, "grad_norm": 0.1530999094247818, "learning_rate": 8.395597019310275e-05, "loss": 0.0329, "step": 47880 }, { "epoch": 0.9569961232564645, "grad_norm": 0.14159969985485077, "learning_rate": 8.39483838876396e-05, "loss": 0.0357, "step": 47890 }, { "epoch": 0.9571959553974662, "grad_norm": 0.1783992201089859, "learning_rate": 8.394079613196389e-05, "loss": 0.0399, "step": 47900 }, { "epoch": 0.9573957875384677, "grad_norm": 0.15411417186260223, "learning_rate": 8.393320692639969e-05, "loss": 0.0327, "step": 47910 }, { "epoch": 0.9575956196794693, "grad_norm": 0.1395197957754135, "learning_rate": 8.392561627127124e-05, "loss": 0.0375, "step": 47920 }, { "epoch": 0.9577954518204708, "grad_norm": 0.1788167506456375, "learning_rate": 8.391802416690279e-05, "loss": 0.0352, "step": 47930 }, { "epoch": 0.9579952839614724, "grad_norm": 0.17412841320037842, "learning_rate": 8.391043061361864e-05, "loss": 0.0284, "step": 47940 }, { "epoch": 0.9581951161024739, "grad_norm": 0.11484990268945694, "learning_rate": 8.390283561174321e-05, "loss": 0.0327, "step": 47950 }, { "epoch": 0.9583949482434755, "grad_norm": 0.12652383744716644, "learning_rate": 8.389523916160092e-05, "loss": 0.0374, "step": 47960 }, { "epoch": 0.958594780384477, "grad_norm": 0.13785384595394135, "learning_rate": 8.388764126351626e-05, "loss": 0.0329, "step": 47970 }, { "epoch": 0.9587946125254786, "grad_norm": 0.14436157047748566, "learning_rate": 8.388004191781383e-05, "loss": 0.0437, "step": 47980 }, { "epoch": 0.9589944446664802, "grad_norm": 0.12023277580738068, "learning_rate": 8.387244112481827e-05, "loss": 0.0449, "step": 47990 }, { "epoch": 0.9591942768074817, "grad_norm": 0.12286742031574249, "learning_rate": 8.386483888485422e-05, "loss": 0.0374, "step": 48000 }, { "epoch": 0.9593941089484833, "grad_norm": 0.1441151201725006, "learning_rate": 8.385723519824649e-05, "loss": 0.0397, "step": 48010 }, { "epoch": 0.9595939410894848, "grad_norm": 0.12587691843509674, "learning_rate": 8.384963006531985e-05, "loss": 0.0383, "step": 48020 }, { "epoch": 0.9597937732304864, "grad_norm": 0.12384304404258728, "learning_rate": 8.38420234863992e-05, "loss": 0.0375, "step": 48030 }, { "epoch": 0.9599936053714879, "grad_norm": 0.1352173089981079, "learning_rate": 8.38344154618095e-05, "loss": 0.0451, "step": 48040 }, { "epoch": 0.9601934375124895, "grad_norm": 0.14939555525779724, "learning_rate": 8.38268059918757e-05, "loss": 0.0357, "step": 48050 }, { "epoch": 0.960393269653491, "grad_norm": 0.13031341135501862, "learning_rate": 8.381919507692291e-05, "loss": 0.0251, "step": 48060 }, { "epoch": 0.9605931017944926, "grad_norm": 0.1584566831588745, "learning_rate": 8.381158271727621e-05, "loss": 0.0411, "step": 48070 }, { "epoch": 0.9607929339354941, "grad_norm": 0.12192699313163757, "learning_rate": 8.380396891326083e-05, "loss": 0.0306, "step": 48080 }, { "epoch": 0.9609927660764958, "grad_norm": 0.1292913556098938, "learning_rate": 8.379635366520199e-05, "loss": 0.0388, "step": 48090 }, { "epoch": 0.9611925982174973, "grad_norm": 0.13434723019599915, "learning_rate": 8.378873697342501e-05, "loss": 0.0364, "step": 48100 }, { "epoch": 0.9613924303584989, "grad_norm": 0.13476142287254333, "learning_rate": 8.378111883825525e-05, "loss": 0.039, "step": 48110 }, { "epoch": 0.9615922624995005, "grad_norm": 0.12104436755180359, "learning_rate": 8.377349926001816e-05, "loss": 0.0322, "step": 48120 }, { "epoch": 0.961792094640502, "grad_norm": 0.15160810947418213, "learning_rate": 8.376587823903923e-05, "loss": 0.0303, "step": 48130 }, { "epoch": 0.9619919267815036, "grad_norm": 0.19021597504615784, "learning_rate": 8.3758255775644e-05, "loss": 0.04, "step": 48140 }, { "epoch": 0.9621917589225051, "grad_norm": 0.13464906811714172, "learning_rate": 8.37506318701581e-05, "loss": 0.0399, "step": 48150 }, { "epoch": 0.9623915910635067, "grad_norm": 0.1068926453590393, "learning_rate": 8.374300652290722e-05, "loss": 0.03, "step": 48160 }, { "epoch": 0.9625914232045082, "grad_norm": 0.12124232202768326, "learning_rate": 8.373537973421708e-05, "loss": 0.0474, "step": 48170 }, { "epoch": 0.9627912553455098, "grad_norm": 0.1661926805973053, "learning_rate": 8.37277515044135e-05, "loss": 0.038, "step": 48180 }, { "epoch": 0.9629910874865113, "grad_norm": 0.1233699694275856, "learning_rate": 8.372012183382233e-05, "loss": 0.0326, "step": 48190 }, { "epoch": 0.9631909196275129, "grad_norm": 0.15328417718410492, "learning_rate": 8.37124907227695e-05, "loss": 0.0307, "step": 48200 }, { "epoch": 0.9633907517685144, "grad_norm": 0.146651029586792, "learning_rate": 8.370485817158101e-05, "loss": 0.0382, "step": 48210 }, { "epoch": 0.963590583909516, "grad_norm": 0.1127186045050621, "learning_rate": 8.369722418058288e-05, "loss": 0.0367, "step": 48220 }, { "epoch": 0.9637904160505175, "grad_norm": 0.17561596632003784, "learning_rate": 8.368958875010123e-05, "loss": 0.0396, "step": 48230 }, { "epoch": 0.9639902481915191, "grad_norm": 0.1687934547662735, "learning_rate": 8.368195188046226e-05, "loss": 0.0322, "step": 48240 }, { "epoch": 0.9641900803325207, "grad_norm": 0.09259683638811111, "learning_rate": 8.367431357199217e-05, "loss": 0.0346, "step": 48250 }, { "epoch": 0.9643899124735222, "grad_norm": 0.11574596911668777, "learning_rate": 8.366667382501728e-05, "loss": 0.0337, "step": 48260 }, { "epoch": 0.9645897446145238, "grad_norm": 0.17181023955345154, "learning_rate": 8.36590326398639e-05, "loss": 0.0374, "step": 48270 }, { "epoch": 0.9647895767555253, "grad_norm": 0.1400812268257141, "learning_rate": 8.36513900168585e-05, "loss": 0.0328, "step": 48280 }, { "epoch": 0.964989408896527, "grad_norm": 0.10746637731790543, "learning_rate": 8.364374595632754e-05, "loss": 0.0307, "step": 48290 }, { "epoch": 0.9651892410375285, "grad_norm": 0.13030163943767548, "learning_rate": 8.363610045859756e-05, "loss": 0.035, "step": 48300 }, { "epoch": 0.9653890731785301, "grad_norm": 0.14292940497398376, "learning_rate": 8.362845352399515e-05, "loss": 0.0371, "step": 48310 }, { "epoch": 0.9655889053195316, "grad_norm": 0.14223450422286987, "learning_rate": 8.3620805152847e-05, "loss": 0.0354, "step": 48320 }, { "epoch": 0.9657887374605332, "grad_norm": 0.12839816510677338, "learning_rate": 8.36131553454798e-05, "loss": 0.0328, "step": 48330 }, { "epoch": 0.9659885696015347, "grad_norm": 0.14309370517730713, "learning_rate": 8.360550410222036e-05, "loss": 0.0401, "step": 48340 }, { "epoch": 0.9661884017425363, "grad_norm": 0.1035400778055191, "learning_rate": 8.359785142339551e-05, "loss": 0.0351, "step": 48350 }, { "epoch": 0.9663882338835378, "grad_norm": 0.15125642716884613, "learning_rate": 8.359019730933218e-05, "loss": 0.0345, "step": 48360 }, { "epoch": 0.9665880660245394, "grad_norm": 0.11706200987100601, "learning_rate": 8.358254176035733e-05, "loss": 0.039, "step": 48370 }, { "epoch": 0.9667878981655409, "grad_norm": 0.1370156705379486, "learning_rate": 8.357488477679798e-05, "loss": 0.0437, "step": 48380 }, { "epoch": 0.9669877303065425, "grad_norm": 0.11013436317443848, "learning_rate": 8.356722635898125e-05, "loss": 0.0336, "step": 48390 }, { "epoch": 0.9671875624475441, "grad_norm": 0.10987289994955063, "learning_rate": 8.355956650723425e-05, "loss": 0.0355, "step": 48400 }, { "epoch": 0.9673873945885456, "grad_norm": 0.12367033958435059, "learning_rate": 8.355190522188423e-05, "loss": 0.0375, "step": 48410 }, { "epoch": 0.9675872267295472, "grad_norm": 0.13838078081607819, "learning_rate": 8.354424250325845e-05, "loss": 0.0398, "step": 48420 }, { "epoch": 0.9677870588705487, "grad_norm": 0.11916574090719223, "learning_rate": 8.353657835168426e-05, "loss": 0.0405, "step": 48430 }, { "epoch": 0.9679868910115503, "grad_norm": 0.10289496183395386, "learning_rate": 8.352891276748905e-05, "loss": 0.0348, "step": 48440 }, { "epoch": 0.9681867231525518, "grad_norm": 0.15328390896320343, "learning_rate": 8.352124575100028e-05, "loss": 0.0373, "step": 48450 }, { "epoch": 0.9683865552935534, "grad_norm": 0.1319556087255478, "learning_rate": 8.351357730254547e-05, "loss": 0.0352, "step": 48460 }, { "epoch": 0.9685863874345549, "grad_norm": 0.14959323406219482, "learning_rate": 8.350590742245221e-05, "loss": 0.0359, "step": 48470 }, { "epoch": 0.9687862195755566, "grad_norm": 0.14250609278678894, "learning_rate": 8.349823611104815e-05, "loss": 0.0399, "step": 48480 }, { "epoch": 0.968986051716558, "grad_norm": 0.12700213491916656, "learning_rate": 8.349056336866096e-05, "loss": 0.0392, "step": 48490 }, { "epoch": 0.9691858838575597, "grad_norm": 0.10496813803911209, "learning_rate": 8.348288919561842e-05, "loss": 0.0304, "step": 48500 }, { "epoch": 0.9693857159985612, "grad_norm": 0.1332598626613617, "learning_rate": 8.347521359224839e-05, "loss": 0.0357, "step": 48510 }, { "epoch": 0.9695855481395628, "grad_norm": 0.11128754168748856, "learning_rate": 8.346753655887871e-05, "loss": 0.0375, "step": 48520 }, { "epoch": 0.9697853802805644, "grad_norm": 0.12120775133371353, "learning_rate": 8.345985809583736e-05, "loss": 0.0309, "step": 48530 }, { "epoch": 0.9699852124215659, "grad_norm": 0.10726339370012283, "learning_rate": 8.345217820345233e-05, "loss": 0.0365, "step": 48540 }, { "epoch": 0.9701850445625675, "grad_norm": 0.13234420120716095, "learning_rate": 8.344449688205172e-05, "loss": 0.0316, "step": 48550 }, { "epoch": 0.970384876703569, "grad_norm": 0.14387807250022888, "learning_rate": 8.343681413196362e-05, "loss": 0.0355, "step": 48560 }, { "epoch": 0.9705847088445706, "grad_norm": 0.13527150452136993, "learning_rate": 8.342912995351627e-05, "loss": 0.0333, "step": 48570 }, { "epoch": 0.9707845409855721, "grad_norm": 0.13112446665763855, "learning_rate": 8.342144434703788e-05, "loss": 0.0335, "step": 48580 }, { "epoch": 0.9709843731265737, "grad_norm": 0.1142105683684349, "learning_rate": 8.341375731285679e-05, "loss": 0.0381, "step": 48590 }, { "epoch": 0.9711842052675752, "grad_norm": 0.15211226046085358, "learning_rate": 8.340606885130138e-05, "loss": 0.036, "step": 48600 }, { "epoch": 0.9713840374085768, "grad_norm": 0.1460665613412857, "learning_rate": 8.339837896270006e-05, "loss": 0.0347, "step": 48610 }, { "epoch": 0.9715838695495783, "grad_norm": 0.10801731050014496, "learning_rate": 8.339068764738136e-05, "loss": 0.0332, "step": 48620 }, { "epoch": 0.9717837016905799, "grad_norm": 0.1061801165342331, "learning_rate": 8.338299490567383e-05, "loss": 0.0434, "step": 48630 }, { "epoch": 0.9719835338315814, "grad_norm": 0.1350664496421814, "learning_rate": 8.337530073790608e-05, "loss": 0.0398, "step": 48640 }, { "epoch": 0.972183365972583, "grad_norm": 0.10461978614330292, "learning_rate": 8.336760514440678e-05, "loss": 0.0349, "step": 48650 }, { "epoch": 0.9723831981135845, "grad_norm": 0.11921697854995728, "learning_rate": 8.33599081255047e-05, "loss": 0.0343, "step": 48660 }, { "epoch": 0.9725830302545861, "grad_norm": 0.1393112987279892, "learning_rate": 8.335220968152862e-05, "loss": 0.0341, "step": 48670 }, { "epoch": 0.9727828623955878, "grad_norm": 0.11390114575624466, "learning_rate": 8.334450981280741e-05, "loss": 0.0329, "step": 48680 }, { "epoch": 0.9729826945365893, "grad_norm": 0.12629836797714233, "learning_rate": 8.333680851967e-05, "loss": 0.0306, "step": 48690 }, { "epoch": 0.9731825266775909, "grad_norm": 0.1353786587715149, "learning_rate": 8.332910580244538e-05, "loss": 0.0377, "step": 48700 }, { "epoch": 0.9733823588185924, "grad_norm": 0.1581028550863266, "learning_rate": 8.332140166146258e-05, "loss": 0.0411, "step": 48710 }, { "epoch": 0.973582190959594, "grad_norm": 0.12932509183883667, "learning_rate": 8.331369609705072e-05, "loss": 0.0397, "step": 48720 }, { "epoch": 0.9737820231005955, "grad_norm": 0.15415504574775696, "learning_rate": 8.330598910953896e-05, "loss": 0.0317, "step": 48730 }, { "epoch": 0.9739818552415971, "grad_norm": 0.11468971520662308, "learning_rate": 8.329828069925652e-05, "loss": 0.0314, "step": 48740 }, { "epoch": 0.9741816873825986, "grad_norm": 0.11963459849357605, "learning_rate": 8.329057086653271e-05, "loss": 0.0292, "step": 48750 }, { "epoch": 0.9743815195236002, "grad_norm": 0.16054154932498932, "learning_rate": 8.328285961169687e-05, "loss": 0.0399, "step": 48760 }, { "epoch": 0.9745813516646017, "grad_norm": 0.13728879392147064, "learning_rate": 8.327514693507841e-05, "loss": 0.0336, "step": 48770 }, { "epoch": 0.9747811838056033, "grad_norm": 0.1530868113040924, "learning_rate": 8.32674328370068e-05, "loss": 0.0475, "step": 48780 }, { "epoch": 0.9749810159466048, "grad_norm": 0.16526387631893158, "learning_rate": 8.325971731781156e-05, "loss": 0.0355, "step": 48790 }, { "epoch": 0.9751808480876064, "grad_norm": 0.15657584369182587, "learning_rate": 8.325200037782232e-05, "loss": 0.0353, "step": 48800 }, { "epoch": 0.975380680228608, "grad_norm": 0.17163127660751343, "learning_rate": 8.32442820173687e-05, "loss": 0.0364, "step": 48810 }, { "epoch": 0.9755805123696095, "grad_norm": 0.13971398770809174, "learning_rate": 8.323656223678042e-05, "loss": 0.0377, "step": 48820 }, { "epoch": 0.9757803445106111, "grad_norm": 0.1435053050518036, "learning_rate": 8.322884103638726e-05, "loss": 0.0376, "step": 48830 }, { "epoch": 0.9759801766516126, "grad_norm": 0.17370198667049408, "learning_rate": 8.322111841651906e-05, "loss": 0.0519, "step": 48840 }, { "epoch": 0.9761800087926142, "grad_norm": 0.13016368448734283, "learning_rate": 8.32133943775057e-05, "loss": 0.0326, "step": 48850 }, { "epoch": 0.9763798409336157, "grad_norm": 0.12626053392887115, "learning_rate": 8.320566891967716e-05, "loss": 0.0309, "step": 48860 }, { "epoch": 0.9765796730746173, "grad_norm": 0.11370877176523209, "learning_rate": 8.319794204336344e-05, "loss": 0.0314, "step": 48870 }, { "epoch": 0.9767795052156188, "grad_norm": 0.1190837174654007, "learning_rate": 8.319021374889462e-05, "loss": 0.0328, "step": 48880 }, { "epoch": 0.9769793373566205, "grad_norm": 0.12499864399433136, "learning_rate": 8.318248403660084e-05, "loss": 0.0365, "step": 48890 }, { "epoch": 0.977179169497622, "grad_norm": 0.1360265165567398, "learning_rate": 8.31747529068123e-05, "loss": 0.0454, "step": 48900 }, { "epoch": 0.9773790016386236, "grad_norm": 0.13581493496894836, "learning_rate": 8.316702035985926e-05, "loss": 0.0345, "step": 48910 }, { "epoch": 0.9775788337796251, "grad_norm": 0.15672381222248077, "learning_rate": 8.315928639607204e-05, "loss": 0.0308, "step": 48920 }, { "epoch": 0.9777786659206267, "grad_norm": 0.15138660371303558, "learning_rate": 8.315155101578102e-05, "loss": 0.0456, "step": 48930 }, { "epoch": 0.9779784980616283, "grad_norm": 0.1290164291858673, "learning_rate": 8.314381421931666e-05, "loss": 0.0411, "step": 48940 }, { "epoch": 0.9781783302026298, "grad_norm": 0.1525397151708603, "learning_rate": 8.313607600700941e-05, "loss": 0.0382, "step": 48950 }, { "epoch": 0.9783781623436314, "grad_norm": 0.19502374529838562, "learning_rate": 8.312833637918989e-05, "loss": 0.0332, "step": 48960 }, { "epoch": 0.9785779944846329, "grad_norm": 0.14392247796058655, "learning_rate": 8.312059533618869e-05, "loss": 0.0361, "step": 48970 }, { "epoch": 0.9787778266256345, "grad_norm": 0.17497192323207855, "learning_rate": 8.31128528783365e-05, "loss": 0.0311, "step": 48980 }, { "epoch": 0.978977658766636, "grad_norm": 0.16747568547725677, "learning_rate": 8.310510900596407e-05, "loss": 0.0306, "step": 48990 }, { "epoch": 0.9791774909076376, "grad_norm": 0.12797802686691284, "learning_rate": 8.309736371940219e-05, "loss": 0.031, "step": 49000 }, { "epoch": 0.9793773230486391, "grad_norm": 0.15808390080928802, "learning_rate": 8.308961701898174e-05, "loss": 0.0321, "step": 49010 }, { "epoch": 0.9795771551896407, "grad_norm": 0.17924383282661438, "learning_rate": 8.308186890503366e-05, "loss": 0.0324, "step": 49020 }, { "epoch": 0.9797769873306422, "grad_norm": 0.1214575320482254, "learning_rate": 8.307411937788888e-05, "loss": 0.0419, "step": 49030 }, { "epoch": 0.9799768194716438, "grad_norm": 0.12933585047721863, "learning_rate": 8.30663684378785e-05, "loss": 0.0324, "step": 49040 }, { "epoch": 0.9801766516126453, "grad_norm": 0.15705975890159607, "learning_rate": 8.30586160853336e-05, "loss": 0.0409, "step": 49050 }, { "epoch": 0.9803764837536469, "grad_norm": 0.13337908685207367, "learning_rate": 8.305086232058534e-05, "loss": 0.0455, "step": 49060 }, { "epoch": 0.9805763158946484, "grad_norm": 0.15966454148292542, "learning_rate": 8.304310714396498e-05, "loss": 0.0345, "step": 49070 }, { "epoch": 0.98077614803565, "grad_norm": 0.16364453732967377, "learning_rate": 8.303535055580377e-05, "loss": 0.0423, "step": 49080 }, { "epoch": 0.9809759801766517, "grad_norm": 0.11091295629739761, "learning_rate": 8.302759255643307e-05, "loss": 0.0369, "step": 49090 }, { "epoch": 0.9811758123176532, "grad_norm": 0.12603802978992462, "learning_rate": 8.301983314618428e-05, "loss": 0.0326, "step": 49100 }, { "epoch": 0.9813756444586548, "grad_norm": 0.14165590703487396, "learning_rate": 8.301207232538891e-05, "loss": 0.0317, "step": 49110 }, { "epoch": 0.9815754765996563, "grad_norm": 0.13119067251682281, "learning_rate": 8.300431009437843e-05, "loss": 0.0324, "step": 49120 }, { "epoch": 0.9817753087406579, "grad_norm": 0.13415966928005219, "learning_rate": 8.299654645348448e-05, "loss": 0.0308, "step": 49130 }, { "epoch": 0.9819751408816594, "grad_norm": 0.16251692175865173, "learning_rate": 8.298878140303865e-05, "loss": 0.0327, "step": 49140 }, { "epoch": 0.982174973022661, "grad_norm": 0.12722457945346832, "learning_rate": 8.29810149433727e-05, "loss": 0.0331, "step": 49150 }, { "epoch": 0.9823748051636625, "grad_norm": 0.11369764804840088, "learning_rate": 8.297324707481838e-05, "loss": 0.0274, "step": 49160 }, { "epoch": 0.9825746373046641, "grad_norm": 0.15272346138954163, "learning_rate": 8.29654777977075e-05, "loss": 0.0439, "step": 49170 }, { "epoch": 0.9827744694456656, "grad_norm": 0.1275649219751358, "learning_rate": 8.295770711237199e-05, "loss": 0.0431, "step": 49180 }, { "epoch": 0.9829743015866672, "grad_norm": 0.15468528866767883, "learning_rate": 8.294993501914377e-05, "loss": 0.0406, "step": 49190 }, { "epoch": 0.9831741337276687, "grad_norm": 0.16502295434474945, "learning_rate": 8.294216151835485e-05, "loss": 0.0412, "step": 49200 }, { "epoch": 0.9833739658686703, "grad_norm": 0.14929690957069397, "learning_rate": 8.29343866103373e-05, "loss": 0.038, "step": 49210 }, { "epoch": 0.9835737980096719, "grad_norm": 0.12995916604995728, "learning_rate": 8.292661029542327e-05, "loss": 0.0357, "step": 49220 }, { "epoch": 0.9837736301506734, "grad_norm": 0.24090567231178284, "learning_rate": 8.291883257394493e-05, "loss": 0.038, "step": 49230 }, { "epoch": 0.983973462291675, "grad_norm": 0.15983250737190247, "learning_rate": 8.291105344623456e-05, "loss": 0.0444, "step": 49240 }, { "epoch": 0.9841732944326765, "grad_norm": 0.1271904855966568, "learning_rate": 8.290327291262441e-05, "loss": 0.0328, "step": 49250 }, { "epoch": 0.9843731265736781, "grad_norm": 0.12469276040792465, "learning_rate": 8.289549097344691e-05, "loss": 0.0428, "step": 49260 }, { "epoch": 0.9845729587146796, "grad_norm": 0.13137324154376984, "learning_rate": 8.288770762903446e-05, "loss": 0.0305, "step": 49270 }, { "epoch": 0.9847727908556813, "grad_norm": 0.1904536634683609, "learning_rate": 8.287992287971955e-05, "loss": 0.0316, "step": 49280 }, { "epoch": 0.9849726229966828, "grad_norm": 0.10700008273124695, "learning_rate": 8.287213672583474e-05, "loss": 0.0349, "step": 49290 }, { "epoch": 0.9851724551376844, "grad_norm": 0.16311439871788025, "learning_rate": 8.286434916771265e-05, "loss": 0.0322, "step": 49300 }, { "epoch": 0.9853722872786859, "grad_norm": 0.17364342510700226, "learning_rate": 8.285656020568592e-05, "loss": 0.0329, "step": 49310 }, { "epoch": 0.9855721194196875, "grad_norm": 0.11625998467206955, "learning_rate": 8.284876984008731e-05, "loss": 0.0353, "step": 49320 }, { "epoch": 0.985771951560689, "grad_norm": 0.13459591567516327, "learning_rate": 8.28409780712496e-05, "loss": 0.0381, "step": 49330 }, { "epoch": 0.9859717837016906, "grad_norm": 0.12943977117538452, "learning_rate": 8.283318489950562e-05, "loss": 0.039, "step": 49340 }, { "epoch": 0.9861716158426921, "grad_norm": 0.14793483912944794, "learning_rate": 8.282539032518832e-05, "loss": 0.0401, "step": 49350 }, { "epoch": 0.9863714479836937, "grad_norm": 0.13118761777877808, "learning_rate": 8.281759434863063e-05, "loss": 0.0289, "step": 49360 }, { "epoch": 0.9865712801246953, "grad_norm": 0.11411511898040771, "learning_rate": 8.28097969701656e-05, "loss": 0.0394, "step": 49370 }, { "epoch": 0.9867711122656968, "grad_norm": 0.1509418934583664, "learning_rate": 8.280199819012632e-05, "loss": 0.0447, "step": 49380 }, { "epoch": 0.9869709444066984, "grad_norm": 0.12031495571136475, "learning_rate": 8.279419800884594e-05, "loss": 0.0416, "step": 49390 }, { "epoch": 0.9871707765476999, "grad_norm": 0.10569722205400467, "learning_rate": 8.278639642665766e-05, "loss": 0.0316, "step": 49400 }, { "epoch": 0.9873706086887015, "grad_norm": 0.1179196909070015, "learning_rate": 8.277859344389476e-05, "loss": 0.029, "step": 49410 }, { "epoch": 0.987570440829703, "grad_norm": 0.16470928490161896, "learning_rate": 8.277078906089056e-05, "loss": 0.0314, "step": 49420 }, { "epoch": 0.9877702729707046, "grad_norm": 0.12074147909879684, "learning_rate": 8.276298327797845e-05, "loss": 0.0398, "step": 49430 }, { "epoch": 0.9879701051117061, "grad_norm": 0.12209653109312057, "learning_rate": 8.275517609549188e-05, "loss": 0.0335, "step": 49440 }, { "epoch": 0.9881699372527077, "grad_norm": 0.14624275267124176, "learning_rate": 8.274736751376437e-05, "loss": 0.0438, "step": 49450 }, { "epoch": 0.9883697693937092, "grad_norm": 0.14270256459712982, "learning_rate": 8.273955753312948e-05, "loss": 0.0284, "step": 49460 }, { "epoch": 0.9885696015347109, "grad_norm": 0.1660398691892624, "learning_rate": 8.273174615392083e-05, "loss": 0.031, "step": 49470 }, { "epoch": 0.9887694336757124, "grad_norm": 0.14756394922733307, "learning_rate": 8.272393337647211e-05, "loss": 0.0308, "step": 49480 }, { "epoch": 0.988969265816714, "grad_norm": 0.13520249724388123, "learning_rate": 8.271611920111709e-05, "loss": 0.0374, "step": 49490 }, { "epoch": 0.9891690979577156, "grad_norm": 0.20458482205867767, "learning_rate": 8.270830362818955e-05, "loss": 0.0356, "step": 49500 }, { "epoch": 0.9893689300987171, "grad_norm": 0.13143791258335114, "learning_rate": 8.270048665802336e-05, "loss": 0.028, "step": 49510 }, { "epoch": 0.9895687622397187, "grad_norm": 0.13858681917190552, "learning_rate": 8.269266829095246e-05, "loss": 0.0266, "step": 49520 }, { "epoch": 0.9897685943807202, "grad_norm": 0.1887265145778656, "learning_rate": 8.268484852731082e-05, "loss": 0.0322, "step": 49530 }, { "epoch": 0.9899684265217218, "grad_norm": 0.11832159757614136, "learning_rate": 8.26770273674325e-05, "loss": 0.0324, "step": 49540 }, { "epoch": 0.9901682586627233, "grad_norm": 0.15730524063110352, "learning_rate": 8.26692048116516e-05, "loss": 0.0349, "step": 49550 }, { "epoch": 0.9903680908037249, "grad_norm": 0.13912928104400635, "learning_rate": 8.266138086030228e-05, "loss": 0.0357, "step": 49560 }, { "epoch": 0.9905679229447264, "grad_norm": 0.1349630355834961, "learning_rate": 8.265355551371879e-05, "loss": 0.0358, "step": 49570 }, { "epoch": 0.990767755085728, "grad_norm": 0.1421460658311844, "learning_rate": 8.264572877223538e-05, "loss": 0.0367, "step": 49580 }, { "epoch": 0.9909675872267295, "grad_norm": 0.12750086188316345, "learning_rate": 8.26379006361864e-05, "loss": 0.0351, "step": 49590 }, { "epoch": 0.9911674193677311, "grad_norm": 0.20451630651950836, "learning_rate": 8.263007110590628e-05, "loss": 0.0338, "step": 49600 }, { "epoch": 0.9913672515087326, "grad_norm": 0.09939858317375183, "learning_rate": 8.262224018172945e-05, "loss": 0.033, "step": 49610 }, { "epoch": 0.9915670836497342, "grad_norm": 0.10543514788150787, "learning_rate": 8.261440786399047e-05, "loss": 0.0323, "step": 49620 }, { "epoch": 0.9917669157907358, "grad_norm": 0.10470816493034363, "learning_rate": 8.26065741530239e-05, "loss": 0.0286, "step": 49630 }, { "epoch": 0.9919667479317373, "grad_norm": 0.12519071996212006, "learning_rate": 8.259873904916435e-05, "loss": 0.0298, "step": 49640 }, { "epoch": 0.992166580072739, "grad_norm": 0.12785780429840088, "learning_rate": 8.259090255274659e-05, "loss": 0.0297, "step": 49650 }, { "epoch": 0.9923664122137404, "grad_norm": 0.13506513833999634, "learning_rate": 8.258306466410532e-05, "loss": 0.0406, "step": 49660 }, { "epoch": 0.992566244354742, "grad_norm": 0.1499597132205963, "learning_rate": 8.25752253835754e-05, "loss": 0.0367, "step": 49670 }, { "epoch": 0.9927660764957436, "grad_norm": 0.15892726182937622, "learning_rate": 8.256738471149168e-05, "loss": 0.036, "step": 49680 }, { "epoch": 0.9929659086367452, "grad_norm": 0.16835099458694458, "learning_rate": 8.255954264818913e-05, "loss": 0.0404, "step": 49690 }, { "epoch": 0.9931657407777467, "grad_norm": 0.1593763679265976, "learning_rate": 8.255169919400272e-05, "loss": 0.0367, "step": 49700 }, { "epoch": 0.9933655729187483, "grad_norm": 0.13856779038906097, "learning_rate": 8.254385434926752e-05, "loss": 0.0348, "step": 49710 }, { "epoch": 0.9935654050597498, "grad_norm": 0.11586121469736099, "learning_rate": 8.253600811431865e-05, "loss": 0.0419, "step": 49720 }, { "epoch": 0.9937652372007514, "grad_norm": 0.1090000793337822, "learning_rate": 8.252816048949128e-05, "loss": 0.0274, "step": 49730 }, { "epoch": 0.9939650693417529, "grad_norm": 0.10886836051940918, "learning_rate": 8.252031147512066e-05, "loss": 0.0274, "step": 49740 }, { "epoch": 0.9941649014827545, "grad_norm": 0.14695677161216736, "learning_rate": 8.251246107154207e-05, "loss": 0.0402, "step": 49750 }, { "epoch": 0.994364733623756, "grad_norm": 0.11466488987207413, "learning_rate": 8.250460927909087e-05, "loss": 0.0343, "step": 49760 }, { "epoch": 0.9945645657647576, "grad_norm": 0.1332055777311325, "learning_rate": 8.249675609810247e-05, "loss": 0.0314, "step": 49770 }, { "epoch": 0.9947643979057592, "grad_norm": 0.15183010697364807, "learning_rate": 8.248890152891235e-05, "loss": 0.0397, "step": 49780 }, { "epoch": 0.9949642300467607, "grad_norm": 0.125568687915802, "learning_rate": 8.248104557185605e-05, "loss": 0.0331, "step": 49790 }, { "epoch": 0.9951640621877623, "grad_norm": 0.12404438108205795, "learning_rate": 8.247318822726913e-05, "loss": 0.0355, "step": 49800 }, { "epoch": 0.9953638943287638, "grad_norm": 0.10471286624670029, "learning_rate": 8.246532949548728e-05, "loss": 0.0303, "step": 49810 }, { "epoch": 0.9955637264697654, "grad_norm": 0.1652461588382721, "learning_rate": 8.24574693768462e-05, "loss": 0.0362, "step": 49820 }, { "epoch": 0.9957635586107669, "grad_norm": 0.13917674124240875, "learning_rate": 8.244960787168166e-05, "loss": 0.0314, "step": 49830 }, { "epoch": 0.9959633907517685, "grad_norm": 0.11959408968687057, "learning_rate": 8.244174498032947e-05, "loss": 0.032, "step": 49840 }, { "epoch": 0.99616322289277, "grad_norm": 0.13016101717948914, "learning_rate": 8.243388070312551e-05, "loss": 0.0346, "step": 49850 }, { "epoch": 0.9963630550337716, "grad_norm": 0.15872199833393097, "learning_rate": 8.24260150404058e-05, "loss": 0.0418, "step": 49860 }, { "epoch": 0.9965628871747731, "grad_norm": 0.132293701171875, "learning_rate": 8.241814799250625e-05, "loss": 0.0406, "step": 49870 }, { "epoch": 0.9967627193157748, "grad_norm": 0.12551768124103546, "learning_rate": 8.2410279559763e-05, "loss": 0.039, "step": 49880 }, { "epoch": 0.9969625514567763, "grad_norm": 0.11868654191493988, "learning_rate": 8.240240974251214e-05, "loss": 0.0307, "step": 49890 }, { "epoch": 0.9971623835977779, "grad_norm": 0.1459670215845108, "learning_rate": 8.239453854108987e-05, "loss": 0.0362, "step": 49900 }, { "epoch": 0.9973622157387795, "grad_norm": 0.10992193222045898, "learning_rate": 8.23866659558324e-05, "loss": 0.0365, "step": 49910 }, { "epoch": 0.997562047879781, "grad_norm": 0.1403629630804062, "learning_rate": 8.237879198707608e-05, "loss": 0.0285, "step": 49920 }, { "epoch": 0.9977618800207826, "grad_norm": 0.12762047350406647, "learning_rate": 8.237091663515722e-05, "loss": 0.0329, "step": 49930 }, { "epoch": 0.9979617121617841, "grad_norm": 0.14084893465042114, "learning_rate": 8.23630399004123e-05, "loss": 0.0456, "step": 49940 }, { "epoch": 0.9981615443027857, "grad_norm": 0.14642611145973206, "learning_rate": 8.235516178317775e-05, "loss": 0.0359, "step": 49950 }, { "epoch": 0.9983613764437872, "grad_norm": 0.31293800473213196, "learning_rate": 8.234728228379012e-05, "loss": 0.0342, "step": 49960 }, { "epoch": 0.9985612085847888, "grad_norm": 0.1739325225353241, "learning_rate": 8.233940140258604e-05, "loss": 0.0336, "step": 49970 }, { "epoch": 0.9987610407257903, "grad_norm": 0.1337142139673233, "learning_rate": 8.233151913990212e-05, "loss": 0.0381, "step": 49980 }, { "epoch": 0.9989608728667919, "grad_norm": 0.17129550874233246, "learning_rate": 8.232363549607512e-05, "loss": 0.0341, "step": 49990 }, { "epoch": 0.9991607050077934, "grad_norm": 0.13716205954551697, "learning_rate": 8.231575047144177e-05, "loss": 0.0324, "step": 50000 }, { "epoch": 0.999360537148795, "grad_norm": 0.18481770157814026, "learning_rate": 8.230786406633893e-05, "loss": 0.0503, "step": 50010 }, { "epoch": 0.9995603692897965, "grad_norm": 0.13386130332946777, "learning_rate": 8.229997628110349e-05, "loss": 0.0286, "step": 50020 }, { "epoch": 0.9997602014307981, "grad_norm": 0.1401970088481903, "learning_rate": 8.22920871160724e-05, "loss": 0.0329, "step": 50030 }, { "epoch": 0.9999600335717996, "grad_norm": 0.11907262355089188, "learning_rate": 8.228419657158267e-05, "loss": 0.0403, "step": 50040 }, { "epoch": 1.0001598657128012, "grad_norm": 0.15726254880428314, "learning_rate": 8.227630464797138e-05, "loss": 0.0307, "step": 50050 }, { "epoch": 1.0003596978538027, "grad_norm": 0.1394999772310257, "learning_rate": 8.226841134557564e-05, "loss": 0.0361, "step": 50060 }, { "epoch": 1.0005595299948045, "grad_norm": 0.14068672060966492, "learning_rate": 8.226051666473265e-05, "loss": 0.0313, "step": 50070 }, { "epoch": 1.000759362135806, "grad_norm": 0.13606558740139008, "learning_rate": 8.225262060577965e-05, "loss": 0.0312, "step": 50080 }, { "epoch": 1.0009591942768075, "grad_norm": 0.15045957267284393, "learning_rate": 8.224472316905394e-05, "loss": 0.0321, "step": 50090 }, { "epoch": 1.001159026417809, "grad_norm": 0.11948741972446442, "learning_rate": 8.22368243548929e-05, "loss": 0.0376, "step": 50100 }, { "epoch": 1.0013588585588107, "grad_norm": 0.15387624502182007, "learning_rate": 8.222892416363395e-05, "loss": 0.0293, "step": 50110 }, { "epoch": 1.0015586906998122, "grad_norm": 0.13660329580307007, "learning_rate": 8.222102259561455e-05, "loss": 0.0427, "step": 50120 }, { "epoch": 1.0017585228408137, "grad_norm": 0.11369406431913376, "learning_rate": 8.221311965117226e-05, "loss": 0.0312, "step": 50130 }, { "epoch": 1.0019583549818152, "grad_norm": 0.10739761590957642, "learning_rate": 8.220521533064469e-05, "loss": 0.032, "step": 50140 }, { "epoch": 1.002158187122817, "grad_norm": 0.09174841642379761, "learning_rate": 8.219730963436948e-05, "loss": 0.0332, "step": 50150 }, { "epoch": 1.0023580192638184, "grad_norm": 0.1480918973684311, "learning_rate": 8.218940256268435e-05, "loss": 0.0318, "step": 50160 }, { "epoch": 1.00255785140482, "grad_norm": 0.10086330026388168, "learning_rate": 8.218149411592708e-05, "loss": 0.0307, "step": 50170 }, { "epoch": 1.0027576835458216, "grad_norm": 0.16076841950416565, "learning_rate": 8.21735842944355e-05, "loss": 0.0345, "step": 50180 }, { "epoch": 1.0029575156868231, "grad_norm": 0.12570221722126007, "learning_rate": 8.216567309854749e-05, "loss": 0.0348, "step": 50190 }, { "epoch": 1.0031573478278246, "grad_norm": 0.1375730186700821, "learning_rate": 8.215776052860103e-05, "loss": 0.0366, "step": 50200 }, { "epoch": 1.0033571799688261, "grad_norm": 0.09910040348768234, "learning_rate": 8.21498465849341e-05, "loss": 0.0396, "step": 50210 }, { "epoch": 1.0035570121098278, "grad_norm": 0.14419232308864594, "learning_rate": 8.21419312678848e-05, "loss": 0.0425, "step": 50220 }, { "epoch": 1.0037568442508293, "grad_norm": 0.13419176638126373, "learning_rate": 8.213401457779124e-05, "loss": 0.0374, "step": 50230 }, { "epoch": 1.0039566763918308, "grad_norm": 0.13848060369491577, "learning_rate": 8.212609651499159e-05, "loss": 0.0367, "step": 50240 }, { "epoch": 1.0041565085328323, "grad_norm": 0.16978932917118073, "learning_rate": 8.211817707982414e-05, "loss": 0.033, "step": 50250 }, { "epoch": 1.004356340673834, "grad_norm": 0.13266588747501373, "learning_rate": 8.211025627262717e-05, "loss": 0.034, "step": 50260 }, { "epoch": 1.0045561728148356, "grad_norm": 0.1153545007109642, "learning_rate": 8.210233409373902e-05, "loss": 0.0315, "step": 50270 }, { "epoch": 1.004756004955837, "grad_norm": 0.12546126544475555, "learning_rate": 8.209441054349814e-05, "loss": 0.0401, "step": 50280 }, { "epoch": 1.0049558370968386, "grad_norm": 0.1946445256471634, "learning_rate": 8.2086485622243e-05, "loss": 0.0384, "step": 50290 }, { "epoch": 1.0051556692378403, "grad_norm": 0.11249689757823944, "learning_rate": 8.207855933031214e-05, "loss": 0.0407, "step": 50300 }, { "epoch": 1.0053555013788418, "grad_norm": 0.14553484320640564, "learning_rate": 8.207063166804416e-05, "loss": 0.0346, "step": 50310 }, { "epoch": 1.0055553335198433, "grad_norm": 0.17906633019447327, "learning_rate": 8.206270263577771e-05, "loss": 0.0369, "step": 50320 }, { "epoch": 1.005755165660845, "grad_norm": 0.11772096902132034, "learning_rate": 8.20547722338515e-05, "loss": 0.0348, "step": 50330 }, { "epoch": 1.0059549978018465, "grad_norm": 0.1492198407649994, "learning_rate": 8.204684046260432e-05, "loss": 0.0343, "step": 50340 }, { "epoch": 1.006154829942848, "grad_norm": 0.20575979351997375, "learning_rate": 8.203890732237496e-05, "loss": 0.0289, "step": 50350 }, { "epoch": 1.0063546620838495, "grad_norm": 0.14055414497852325, "learning_rate": 8.203097281350236e-05, "loss": 0.0362, "step": 50360 }, { "epoch": 1.0065544942248512, "grad_norm": 0.14318686723709106, "learning_rate": 8.202303693632544e-05, "loss": 0.0319, "step": 50370 }, { "epoch": 1.0067543263658527, "grad_norm": 0.13054287433624268, "learning_rate": 8.201509969118321e-05, "loss": 0.0301, "step": 50380 }, { "epoch": 1.0069541585068542, "grad_norm": 0.14705199003219604, "learning_rate": 8.200716107841473e-05, "loss": 0.0372, "step": 50390 }, { "epoch": 1.0071539906478557, "grad_norm": 0.1384158730506897, "learning_rate": 8.199922109835913e-05, "loss": 0.0349, "step": 50400 }, { "epoch": 1.0073538227888574, "grad_norm": 0.10967987775802612, "learning_rate": 8.19912797513556e-05, "loss": 0.026, "step": 50410 }, { "epoch": 1.007553654929859, "grad_norm": 0.10967474430799484, "learning_rate": 8.198333703774334e-05, "loss": 0.032, "step": 50420 }, { "epoch": 1.0077534870708604, "grad_norm": 0.11881845444440842, "learning_rate": 8.19753929578617e-05, "loss": 0.0283, "step": 50430 }, { "epoch": 1.007953319211862, "grad_norm": 0.12624309957027435, "learning_rate": 8.196744751205e-05, "loss": 0.0367, "step": 50440 }, { "epoch": 1.0081531513528637, "grad_norm": 0.11824767291545868, "learning_rate": 8.195950070064768e-05, "loss": 0.0358, "step": 50450 }, { "epoch": 1.0083529834938652, "grad_norm": 0.15516211092472076, "learning_rate": 8.19515525239942e-05, "loss": 0.035, "step": 50460 }, { "epoch": 1.0085528156348667, "grad_norm": 0.13645289838314056, "learning_rate": 8.19436029824291e-05, "loss": 0.0311, "step": 50470 }, { "epoch": 1.0087526477758684, "grad_norm": 0.12488258630037308, "learning_rate": 8.193565207629195e-05, "loss": 0.0342, "step": 50480 }, { "epoch": 1.0089524799168699, "grad_norm": 0.17575614154338837, "learning_rate": 8.192769980592242e-05, "loss": 0.0447, "step": 50490 }, { "epoch": 1.0091523120578714, "grad_norm": 0.13229908049106598, "learning_rate": 8.19197461716602e-05, "loss": 0.0325, "step": 50500 }, { "epoch": 1.0093521441988729, "grad_norm": 0.11915569007396698, "learning_rate": 8.191179117384505e-05, "loss": 0.0352, "step": 50510 }, { "epoch": 1.0095519763398746, "grad_norm": 0.1389225721359253, "learning_rate": 8.190383481281682e-05, "loss": 0.0336, "step": 50520 }, { "epoch": 1.009751808480876, "grad_norm": 0.1511250138282776, "learning_rate": 8.189587708891537e-05, "loss": 0.0389, "step": 50530 }, { "epoch": 1.0099516406218776, "grad_norm": 0.1403825730085373, "learning_rate": 8.188791800248066e-05, "loss": 0.0411, "step": 50540 }, { "epoch": 1.010151472762879, "grad_norm": 0.1208038181066513, "learning_rate": 8.187995755385268e-05, "loss": 0.0355, "step": 50550 }, { "epoch": 1.0103513049038808, "grad_norm": 0.1357601284980774, "learning_rate": 8.187199574337144e-05, "loss": 0.0371, "step": 50560 }, { "epoch": 1.0105511370448823, "grad_norm": 0.12031099945306778, "learning_rate": 8.186403257137711e-05, "loss": 0.0285, "step": 50570 }, { "epoch": 1.0107509691858838, "grad_norm": 0.11755736172199249, "learning_rate": 8.185606803820987e-05, "loss": 0.0363, "step": 50580 }, { "epoch": 1.0109508013268853, "grad_norm": 0.14915230870246887, "learning_rate": 8.18481021442099e-05, "loss": 0.0292, "step": 50590 }, { "epoch": 1.011150633467887, "grad_norm": 0.12727662920951843, "learning_rate": 8.184013488971751e-05, "loss": 0.036, "step": 50600 }, { "epoch": 1.0113504656088885, "grad_norm": 0.15168598294258118, "learning_rate": 8.183216627507307e-05, "loss": 0.0414, "step": 50610 }, { "epoch": 1.01155029774989, "grad_norm": 0.14013981819152832, "learning_rate": 8.182419630061692e-05, "loss": 0.0323, "step": 50620 }, { "epoch": 1.0117501298908917, "grad_norm": 0.13843578100204468, "learning_rate": 8.181622496668961e-05, "loss": 0.0362, "step": 50630 }, { "epoch": 1.0119499620318932, "grad_norm": 0.1316842883825302, "learning_rate": 8.180825227363163e-05, "loss": 0.0302, "step": 50640 }, { "epoch": 1.0121497941728947, "grad_norm": 0.1596851944923401, "learning_rate": 8.180027822178351e-05, "loss": 0.0403, "step": 50650 }, { "epoch": 1.0123496263138962, "grad_norm": 0.12470658123493195, "learning_rate": 8.179230281148593e-05, "loss": 0.0315, "step": 50660 }, { "epoch": 1.012549458454898, "grad_norm": 0.11319644749164581, "learning_rate": 8.178432604307958e-05, "loss": 0.0271, "step": 50670 }, { "epoch": 1.0127492905958995, "grad_norm": 0.1593552827835083, "learning_rate": 8.177634791690523e-05, "loss": 0.0357, "step": 50680 }, { "epoch": 1.012949122736901, "grad_norm": 0.11587166041135788, "learning_rate": 8.176836843330365e-05, "loss": 0.0345, "step": 50690 }, { "epoch": 1.0131489548779025, "grad_norm": 0.10448014736175537, "learning_rate": 8.176038759261573e-05, "loss": 0.0413, "step": 50700 }, { "epoch": 1.0133487870189042, "grad_norm": 0.20102307200431824, "learning_rate": 8.17524053951824e-05, "loss": 0.0335, "step": 50710 }, { "epoch": 1.0135486191599057, "grad_norm": 0.13766957819461823, "learning_rate": 8.174442184134466e-05, "loss": 0.0312, "step": 50720 }, { "epoch": 1.0137484513009072, "grad_norm": 0.1460292488336563, "learning_rate": 8.173643693144353e-05, "loss": 0.0387, "step": 50730 }, { "epoch": 1.013948283441909, "grad_norm": 0.12769699096679688, "learning_rate": 8.17284506658201e-05, "loss": 0.0435, "step": 50740 }, { "epoch": 1.0141481155829104, "grad_norm": 0.12447631359100342, "learning_rate": 8.172046304481557e-05, "loss": 0.0313, "step": 50750 }, { "epoch": 1.014347947723912, "grad_norm": 0.14868544042110443, "learning_rate": 8.17124740687711e-05, "loss": 0.0395, "step": 50760 }, { "epoch": 1.0145477798649134, "grad_norm": 0.10990548878908157, "learning_rate": 8.170448373802802e-05, "loss": 0.0363, "step": 50770 }, { "epoch": 1.0147476120059151, "grad_norm": 0.1123589500784874, "learning_rate": 8.169649205292762e-05, "loss": 0.0276, "step": 50780 }, { "epoch": 1.0149474441469166, "grad_norm": 0.12612994015216827, "learning_rate": 8.168849901381133e-05, "loss": 0.0282, "step": 50790 }, { "epoch": 1.0151472762879181, "grad_norm": 0.10215773433446884, "learning_rate": 8.168050462102055e-05, "loss": 0.0326, "step": 50800 }, { "epoch": 1.0153471084289196, "grad_norm": 0.13661108911037445, "learning_rate": 8.167250887489682e-05, "loss": 0.0298, "step": 50810 }, { "epoch": 1.0155469405699213, "grad_norm": 0.10159146040678024, "learning_rate": 8.166451177578172e-05, "loss": 0.0338, "step": 50820 }, { "epoch": 1.0157467727109228, "grad_norm": 0.14347833395004272, "learning_rate": 8.16565133240168e-05, "loss": 0.0359, "step": 50830 }, { "epoch": 1.0159466048519243, "grad_norm": 0.1161661446094513, "learning_rate": 8.164851351994383e-05, "loss": 0.0441, "step": 50840 }, { "epoch": 1.0161464369929258, "grad_norm": 0.12872985005378723, "learning_rate": 8.164051236390447e-05, "loss": 0.0362, "step": 50850 }, { "epoch": 1.0163462691339276, "grad_norm": 0.14247651398181915, "learning_rate": 8.163250985624057e-05, "loss": 0.0357, "step": 50860 }, { "epoch": 1.016546101274929, "grad_norm": 0.14958861470222473, "learning_rate": 8.162450599729394e-05, "loss": 0.0377, "step": 50870 }, { "epoch": 1.0167459334159306, "grad_norm": 0.14976489543914795, "learning_rate": 8.161650078740652e-05, "loss": 0.0392, "step": 50880 }, { "epoch": 1.0169457655569323, "grad_norm": 0.14611364901065826, "learning_rate": 8.160849422692028e-05, "loss": 0.0391, "step": 50890 }, { "epoch": 1.0171455976979338, "grad_norm": 0.15767860412597656, "learning_rate": 8.160048631617722e-05, "loss": 0.0371, "step": 50900 }, { "epoch": 1.0173454298389353, "grad_norm": 0.14318867027759552, "learning_rate": 8.159247705551945e-05, "loss": 0.0453, "step": 50910 }, { "epoch": 1.0175452619799368, "grad_norm": 0.13394691050052643, "learning_rate": 8.158446644528907e-05, "loss": 0.0365, "step": 50920 }, { "epoch": 1.0177450941209385, "grad_norm": 0.15129084885120392, "learning_rate": 8.157645448582832e-05, "loss": 0.0368, "step": 50930 }, { "epoch": 1.01794492626194, "grad_norm": 0.20077021420001984, "learning_rate": 8.156844117747945e-05, "loss": 0.0285, "step": 50940 }, { "epoch": 1.0181447584029415, "grad_norm": 0.1313750147819519, "learning_rate": 8.156042652058476e-05, "loss": 0.0314, "step": 50950 }, { "epoch": 1.018344590543943, "grad_norm": 0.14926087856292725, "learning_rate": 8.155241051548662e-05, "loss": 0.0334, "step": 50960 }, { "epoch": 1.0185444226849447, "grad_norm": 0.16191616654396057, "learning_rate": 8.154439316252748e-05, "loss": 0.0353, "step": 50970 }, { "epoch": 1.0187442548259462, "grad_norm": 0.16706489026546478, "learning_rate": 8.15363744620498e-05, "loss": 0.0352, "step": 50980 }, { "epoch": 1.0189440869669477, "grad_norm": 0.1593637764453888, "learning_rate": 8.152835441439614e-05, "loss": 0.039, "step": 50990 }, { "epoch": 1.0191439191079492, "grad_norm": 0.14191262423992157, "learning_rate": 8.152033301990911e-05, "loss": 0.0329, "step": 51000 }, { "epoch": 1.019343751248951, "grad_norm": 0.15142226219177246, "learning_rate": 8.151231027893135e-05, "loss": 0.0497, "step": 51010 }, { "epoch": 1.0195435833899524, "grad_norm": 0.1376408338546753, "learning_rate": 8.150428619180559e-05, "loss": 0.0359, "step": 51020 }, { "epoch": 1.019743415530954, "grad_norm": 0.14592444896697998, "learning_rate": 8.149626075887458e-05, "loss": 0.0453, "step": 51030 }, { "epoch": 1.0199432476719557, "grad_norm": 0.11216966807842255, "learning_rate": 8.148823398048119e-05, "loss": 0.0285, "step": 51040 }, { "epoch": 1.0201430798129572, "grad_norm": 0.1435486078262329, "learning_rate": 8.14802058569683e-05, "loss": 0.0429, "step": 51050 }, { "epoch": 1.0203429119539587, "grad_norm": 0.12594261765480042, "learning_rate": 8.147217638867882e-05, "loss": 0.0424, "step": 51060 }, { "epoch": 1.0205427440949602, "grad_norm": 0.13204555213451385, "learning_rate": 8.146414557595578e-05, "loss": 0.0372, "step": 51070 }, { "epoch": 1.0207425762359619, "grad_norm": 0.1250275820493698, "learning_rate": 8.145611341914228e-05, "loss": 0.0351, "step": 51080 }, { "epoch": 1.0209424083769634, "grad_norm": 0.12058275938034058, "learning_rate": 8.144807991858137e-05, "loss": 0.0337, "step": 51090 }, { "epoch": 1.0211422405179649, "grad_norm": 0.13978075981140137, "learning_rate": 8.144004507461627e-05, "loss": 0.0337, "step": 51100 }, { "epoch": 1.0213420726589664, "grad_norm": 0.12810717523097992, "learning_rate": 8.14320088875902e-05, "loss": 0.039, "step": 51110 }, { "epoch": 1.021541904799968, "grad_norm": 0.13456043601036072, "learning_rate": 8.142397135784644e-05, "loss": 0.0385, "step": 51120 }, { "epoch": 1.0217417369409696, "grad_norm": 0.1458367258310318, "learning_rate": 8.141593248572837e-05, "loss": 0.0372, "step": 51130 }, { "epoch": 1.021941569081971, "grad_norm": 0.14291560649871826, "learning_rate": 8.140789227157935e-05, "loss": 0.0425, "step": 51140 }, { "epoch": 1.0221414012229728, "grad_norm": 0.12515468895435333, "learning_rate": 8.13998507157429e-05, "loss": 0.0397, "step": 51150 }, { "epoch": 1.0223412333639743, "grad_norm": 0.1340639740228653, "learning_rate": 8.139180781856248e-05, "loss": 0.0339, "step": 51160 }, { "epoch": 1.0225410655049758, "grad_norm": 0.1246396079659462, "learning_rate": 8.138376358038173e-05, "loss": 0.0364, "step": 51170 }, { "epoch": 1.0227408976459773, "grad_norm": 0.13031695783138275, "learning_rate": 8.137571800154425e-05, "loss": 0.0346, "step": 51180 }, { "epoch": 1.022940729786979, "grad_norm": 0.15811388194561005, "learning_rate": 8.136767108239371e-05, "loss": 0.0411, "step": 51190 }, { "epoch": 1.0231405619279805, "grad_norm": 0.14686106145381927, "learning_rate": 8.135962282327392e-05, "loss": 0.0327, "step": 51200 }, { "epoch": 1.023340394068982, "grad_norm": 0.12904205918312073, "learning_rate": 8.135157322452862e-05, "loss": 0.0402, "step": 51210 }, { "epoch": 1.0235402262099835, "grad_norm": 0.13361158967018127, "learning_rate": 8.134352228650172e-05, "loss": 0.0279, "step": 51220 }, { "epoch": 1.0237400583509852, "grad_norm": 0.16929848492145538, "learning_rate": 8.133547000953714e-05, "loss": 0.034, "step": 51230 }, { "epoch": 1.0239398904919867, "grad_norm": 0.16100098192691803, "learning_rate": 8.132741639397883e-05, "loss": 0.0348, "step": 51240 }, { "epoch": 1.0241397226329882, "grad_norm": 0.1274702101945877, "learning_rate": 8.131936144017086e-05, "loss": 0.0431, "step": 51250 }, { "epoch": 1.0243395547739897, "grad_norm": 0.16094718873500824, "learning_rate": 8.131130514845728e-05, "loss": 0.0366, "step": 51260 }, { "epoch": 1.0245393869149915, "grad_norm": 0.17882296442985535, "learning_rate": 8.130324751918228e-05, "loss": 0.0422, "step": 51270 }, { "epoch": 1.024739219055993, "grad_norm": 0.14501450955867767, "learning_rate": 8.129518855269005e-05, "loss": 0.0267, "step": 51280 }, { "epoch": 1.0249390511969945, "grad_norm": 0.14238181710243225, "learning_rate": 8.128712824932485e-05, "loss": 0.0322, "step": 51290 }, { "epoch": 1.0251388833379962, "grad_norm": 0.1333381086587906, "learning_rate": 8.1279066609431e-05, "loss": 0.0316, "step": 51300 }, { "epoch": 1.0253387154789977, "grad_norm": 0.1545577198266983, "learning_rate": 8.127100363335291e-05, "loss": 0.0357, "step": 51310 }, { "epoch": 1.0255385476199992, "grad_norm": 0.13840226829051971, "learning_rate": 8.126293932143499e-05, "loss": 0.0339, "step": 51320 }, { "epoch": 1.0257383797610007, "grad_norm": 0.13436366617679596, "learning_rate": 8.125487367402171e-05, "loss": 0.0298, "step": 51330 }, { "epoch": 1.0259382119020024, "grad_norm": 0.12552665174007416, "learning_rate": 8.124680669145767e-05, "loss": 0.0345, "step": 51340 }, { "epoch": 1.026138044043004, "grad_norm": 0.13423500955104828, "learning_rate": 8.123873837408744e-05, "loss": 0.0283, "step": 51350 }, { "epoch": 1.0263378761840054, "grad_norm": 0.1216304749250412, "learning_rate": 8.123066872225569e-05, "loss": 0.0394, "step": 51360 }, { "epoch": 1.026537708325007, "grad_norm": 0.1435130089521408, "learning_rate": 8.122259773630714e-05, "loss": 0.0356, "step": 51370 }, { "epoch": 1.0267375404660086, "grad_norm": 0.1393914669752121, "learning_rate": 8.121452541658658e-05, "loss": 0.0383, "step": 51380 }, { "epoch": 1.0269373726070101, "grad_norm": 0.13144031167030334, "learning_rate": 8.120645176343884e-05, "loss": 0.033, "step": 51390 }, { "epoch": 1.0271372047480116, "grad_norm": 0.14266689121723175, "learning_rate": 8.11983767772088e-05, "loss": 0.0456, "step": 51400 }, { "epoch": 1.0273370368890131, "grad_norm": 0.13379976153373718, "learning_rate": 8.119030045824143e-05, "loss": 0.034, "step": 51410 }, { "epoch": 1.0275368690300148, "grad_norm": 0.14498992264270782, "learning_rate": 8.118222280688172e-05, "loss": 0.035, "step": 51420 }, { "epoch": 1.0277367011710163, "grad_norm": 0.11990680545568466, "learning_rate": 8.117414382347473e-05, "loss": 0.0333, "step": 51430 }, { "epoch": 1.0279365333120178, "grad_norm": 0.11419719457626343, "learning_rate": 8.116606350836557e-05, "loss": 0.0392, "step": 51440 }, { "epoch": 1.0281363654530196, "grad_norm": 0.17504848539829254, "learning_rate": 8.115798186189945e-05, "loss": 0.034, "step": 51450 }, { "epoch": 1.028336197594021, "grad_norm": 0.10406384617090225, "learning_rate": 8.11498988844216e-05, "loss": 0.0408, "step": 51460 }, { "epoch": 1.0285360297350226, "grad_norm": 0.269871324300766, "learning_rate": 8.114181457627727e-05, "loss": 0.0405, "step": 51470 }, { "epoch": 1.028735861876024, "grad_norm": 0.1200985386967659, "learning_rate": 8.113372893781181e-05, "loss": 0.0307, "step": 51480 }, { "epoch": 1.0289356940170258, "grad_norm": 0.1602369248867035, "learning_rate": 8.11256419693707e-05, "loss": 0.0271, "step": 51490 }, { "epoch": 1.0291355261580273, "grad_norm": 0.12924203276634216, "learning_rate": 8.11175536712993e-05, "loss": 0.0349, "step": 51500 }, { "epoch": 1.0293353582990288, "grad_norm": 0.12018550932407379, "learning_rate": 8.110946404394319e-05, "loss": 0.0318, "step": 51510 }, { "epoch": 1.0295351904400303, "grad_norm": 0.10858150571584702, "learning_rate": 8.110137308764792e-05, "loss": 0.0351, "step": 51520 }, { "epoch": 1.029735022581032, "grad_norm": 0.10452701896429062, "learning_rate": 8.109328080275911e-05, "loss": 0.0295, "step": 51530 }, { "epoch": 1.0299348547220335, "grad_norm": 0.16812576353549957, "learning_rate": 8.108518718962249e-05, "loss": 0.0334, "step": 51540 }, { "epoch": 1.030134686863035, "grad_norm": 0.11777760833501816, "learning_rate": 8.107709224858377e-05, "loss": 0.0313, "step": 51550 }, { "epoch": 1.0303345190040365, "grad_norm": 0.18406322598457336, "learning_rate": 8.106899597998875e-05, "loss": 0.0535, "step": 51560 }, { "epoch": 1.0305343511450382, "grad_norm": 0.15704657137393951, "learning_rate": 8.106089838418332e-05, "loss": 0.0376, "step": 51570 }, { "epoch": 1.0307341832860397, "grad_norm": 0.12684427201747894, "learning_rate": 8.105279946151336e-05, "loss": 0.0298, "step": 51580 }, { "epoch": 1.0309340154270412, "grad_norm": 0.11995691806077957, "learning_rate": 8.104469921232484e-05, "loss": 0.0348, "step": 51590 }, { "epoch": 1.031133847568043, "grad_norm": 0.12262038141489029, "learning_rate": 8.103659763696383e-05, "loss": 0.0327, "step": 51600 }, { "epoch": 1.0313336797090444, "grad_norm": 0.09223317354917526, "learning_rate": 8.102849473577636e-05, "loss": 0.0325, "step": 51610 }, { "epoch": 1.031533511850046, "grad_norm": 0.1299416422843933, "learning_rate": 8.102039050910862e-05, "loss": 0.0307, "step": 51620 }, { "epoch": 1.0317333439910474, "grad_norm": 0.13452783226966858, "learning_rate": 8.101228495730675e-05, "loss": 0.0285, "step": 51630 }, { "epoch": 1.0319331761320492, "grad_norm": 0.11425936222076416, "learning_rate": 8.100417808071707e-05, "loss": 0.0334, "step": 51640 }, { "epoch": 1.0321330082730507, "grad_norm": 0.1542237401008606, "learning_rate": 8.099606987968584e-05, "loss": 0.0281, "step": 51650 }, { "epoch": 1.0323328404140522, "grad_norm": 0.18493179976940155, "learning_rate": 8.098796035455944e-05, "loss": 0.039, "step": 51660 }, { "epoch": 1.0325326725550537, "grad_norm": 0.14383737742900848, "learning_rate": 8.097984950568433e-05, "loss": 0.0388, "step": 51670 }, { "epoch": 1.0327325046960554, "grad_norm": 0.18859754502773285, "learning_rate": 8.097173733340693e-05, "loss": 0.034, "step": 51680 }, { "epoch": 1.0329323368370569, "grad_norm": 0.1651221364736557, "learning_rate": 8.096362383807383e-05, "loss": 0.0442, "step": 51690 }, { "epoch": 1.0331321689780584, "grad_norm": 0.15750664472579956, "learning_rate": 8.095550902003158e-05, "loss": 0.0392, "step": 51700 }, { "epoch": 1.03333200111906, "grad_norm": 0.13060405850410461, "learning_rate": 8.094739287962687e-05, "loss": 0.0395, "step": 51710 }, { "epoch": 1.0335318332600616, "grad_norm": 0.11828839778900146, "learning_rate": 8.093927541720638e-05, "loss": 0.0358, "step": 51720 }, { "epoch": 1.033731665401063, "grad_norm": 0.10769406706094742, "learning_rate": 8.093115663311688e-05, "loss": 0.031, "step": 51730 }, { "epoch": 1.0339314975420646, "grad_norm": 0.16435888409614563, "learning_rate": 8.09230365277052e-05, "loss": 0.0366, "step": 51740 }, { "epoch": 1.0341313296830663, "grad_norm": 0.16374164819717407, "learning_rate": 8.09149151013182e-05, "loss": 0.0321, "step": 51750 }, { "epoch": 1.0343311618240678, "grad_norm": 0.14089442789554596, "learning_rate": 8.090679235430283e-05, "loss": 0.0384, "step": 51760 }, { "epoch": 1.0345309939650693, "grad_norm": 0.18299491703510284, "learning_rate": 8.089866828700605e-05, "loss": 0.0417, "step": 51770 }, { "epoch": 1.0347308261060708, "grad_norm": 0.15535381436347961, "learning_rate": 8.089054289977493e-05, "loss": 0.0324, "step": 51780 }, { "epoch": 1.0349306582470725, "grad_norm": 0.13253715634346008, "learning_rate": 8.088241619295658e-05, "loss": 0.0352, "step": 51790 }, { "epoch": 1.035130490388074, "grad_norm": 0.12359011173248291, "learning_rate": 8.087428816689812e-05, "loss": 0.0393, "step": 51800 }, { "epoch": 1.0353303225290755, "grad_norm": 0.13125328719615936, "learning_rate": 8.086615882194679e-05, "loss": 0.0296, "step": 51810 }, { "epoch": 1.035530154670077, "grad_norm": 0.1250760555267334, "learning_rate": 8.085802815844987e-05, "loss": 0.0422, "step": 51820 }, { "epoch": 1.0357299868110788, "grad_norm": 0.10377144813537598, "learning_rate": 8.084989617675467e-05, "loss": 0.0321, "step": 51830 }, { "epoch": 1.0359298189520803, "grad_norm": 0.1106070727109909, "learning_rate": 8.084176287720858e-05, "loss": 0.0273, "step": 51840 }, { "epoch": 1.0361296510930817, "grad_norm": 0.10576149821281433, "learning_rate": 8.083362826015904e-05, "loss": 0.0335, "step": 51850 }, { "epoch": 1.0363294832340835, "grad_norm": 0.11744862794876099, "learning_rate": 8.082549232595355e-05, "loss": 0.0318, "step": 51860 }, { "epoch": 1.036529315375085, "grad_norm": 0.12979617714881897, "learning_rate": 8.081735507493963e-05, "loss": 0.0337, "step": 51870 }, { "epoch": 1.0367291475160865, "grad_norm": 0.10062521696090698, "learning_rate": 8.080921650746493e-05, "loss": 0.0351, "step": 51880 }, { "epoch": 1.036928979657088, "grad_norm": 0.12100432813167572, "learning_rate": 8.080107662387711e-05, "loss": 0.0384, "step": 51890 }, { "epoch": 1.0371288117980897, "grad_norm": 0.16517868638038635, "learning_rate": 8.079293542452387e-05, "loss": 0.0341, "step": 51900 }, { "epoch": 1.0373286439390912, "grad_norm": 0.13214664161205292, "learning_rate": 8.078479290975301e-05, "loss": 0.0313, "step": 51910 }, { "epoch": 1.0375284760800927, "grad_norm": 0.1059693917632103, "learning_rate": 8.077664907991234e-05, "loss": 0.0322, "step": 51920 }, { "epoch": 1.0377283082210942, "grad_norm": 0.1412762701511383, "learning_rate": 8.076850393534974e-05, "loss": 0.0346, "step": 51930 }, { "epoch": 1.037928140362096, "grad_norm": 0.16241447627544403, "learning_rate": 8.076035747641321e-05, "loss": 0.0437, "step": 51940 }, { "epoch": 1.0381279725030974, "grad_norm": 0.1493857204914093, "learning_rate": 8.075220970345069e-05, "loss": 0.0292, "step": 51950 }, { "epoch": 1.038327804644099, "grad_norm": 0.13778243958950043, "learning_rate": 8.074406061681028e-05, "loss": 0.0301, "step": 51960 }, { "epoch": 1.0385276367851004, "grad_norm": 0.1260698437690735, "learning_rate": 8.073591021684008e-05, "loss": 0.035, "step": 51970 }, { "epoch": 1.0387274689261021, "grad_norm": 0.135930135846138, "learning_rate": 8.072775850388825e-05, "loss": 0.0388, "step": 51980 }, { "epoch": 1.0389273010671036, "grad_norm": 0.1328863501548767, "learning_rate": 8.071960547830304e-05, "loss": 0.0378, "step": 51990 }, { "epoch": 1.0391271332081051, "grad_norm": 0.13674284517765045, "learning_rate": 8.07114511404327e-05, "loss": 0.0371, "step": 52000 }, { "epoch": 1.0393269653491068, "grad_norm": 0.15283742547035217, "learning_rate": 8.070329549062558e-05, "loss": 0.0343, "step": 52010 }, { "epoch": 1.0395267974901083, "grad_norm": 0.1662316471338272, "learning_rate": 8.06951385292301e-05, "loss": 0.033, "step": 52020 }, { "epoch": 1.0397266296311098, "grad_norm": 0.12272404879331589, "learning_rate": 8.068698025659469e-05, "loss": 0.0338, "step": 52030 }, { "epoch": 1.0399264617721113, "grad_norm": 0.11653866618871689, "learning_rate": 8.067882067306785e-05, "loss": 0.0391, "step": 52040 }, { "epoch": 1.040126293913113, "grad_norm": 0.1194702610373497, "learning_rate": 8.067065977899816e-05, "loss": 0.0425, "step": 52050 }, { "epoch": 1.0403261260541146, "grad_norm": 0.1688816100358963, "learning_rate": 8.06624975747342e-05, "loss": 0.0356, "step": 52060 }, { "epoch": 1.040525958195116, "grad_norm": 0.13353456556797028, "learning_rate": 8.065433406062471e-05, "loss": 0.0345, "step": 52070 }, { "epoch": 1.0407257903361176, "grad_norm": 0.14739324152469635, "learning_rate": 8.064616923701836e-05, "loss": 0.031, "step": 52080 }, { "epoch": 1.0409256224771193, "grad_norm": 0.1426180899143219, "learning_rate": 8.063800310426397e-05, "loss": 0.0302, "step": 52090 }, { "epoch": 1.0411254546181208, "grad_norm": 0.14841142296791077, "learning_rate": 8.062983566271037e-05, "loss": 0.0308, "step": 52100 }, { "epoch": 1.0413252867591223, "grad_norm": 0.15914607048034668, "learning_rate": 8.062166691270646e-05, "loss": 0.032, "step": 52110 }, { "epoch": 1.041525118900124, "grad_norm": 0.16440023481845856, "learning_rate": 8.061349685460117e-05, "loss": 0.0384, "step": 52120 }, { "epoch": 1.0417249510411255, "grad_norm": 0.15470783412456512, "learning_rate": 8.060532548874356e-05, "loss": 0.0291, "step": 52130 }, { "epoch": 1.041924783182127, "grad_norm": 0.15676426887512207, "learning_rate": 8.059715281548265e-05, "loss": 0.028, "step": 52140 }, { "epoch": 1.0421246153231285, "grad_norm": 0.11219871044158936, "learning_rate": 8.058897883516759e-05, "loss": 0.0355, "step": 52150 }, { "epoch": 1.0423244474641302, "grad_norm": 0.21334704756736755, "learning_rate": 8.058080354814755e-05, "loss": 0.0379, "step": 52160 }, { "epoch": 1.0425242796051317, "grad_norm": 0.14588584005832672, "learning_rate": 8.057262695477176e-05, "loss": 0.0335, "step": 52170 }, { "epoch": 1.0427241117461332, "grad_norm": 0.14839744567871094, "learning_rate": 8.056444905538951e-05, "loss": 0.0354, "step": 52180 }, { "epoch": 1.0429239438871347, "grad_norm": 0.1273288130760193, "learning_rate": 8.055626985035015e-05, "loss": 0.0393, "step": 52190 }, { "epoch": 1.0431237760281364, "grad_norm": 0.1123337671160698, "learning_rate": 8.054808934000306e-05, "loss": 0.0397, "step": 52200 }, { "epoch": 1.043323608169138, "grad_norm": 0.11826726049184799, "learning_rate": 8.053990752469773e-05, "loss": 0.0303, "step": 52210 }, { "epoch": 1.0435234403101394, "grad_norm": 0.12177567183971405, "learning_rate": 8.053172440478365e-05, "loss": 0.0411, "step": 52220 }, { "epoch": 1.043723272451141, "grad_norm": 0.16817396879196167, "learning_rate": 8.052353998061038e-05, "loss": 0.0352, "step": 52230 }, { "epoch": 1.0439231045921427, "grad_norm": 0.1670517772436142, "learning_rate": 8.051535425252756e-05, "loss": 0.0373, "step": 52240 }, { "epoch": 1.0441229367331442, "grad_norm": 0.15827324986457825, "learning_rate": 8.050716722088488e-05, "loss": 0.0314, "step": 52250 }, { "epoch": 1.0443227688741457, "grad_norm": 0.13945555686950684, "learning_rate": 8.049897888603203e-05, "loss": 0.0353, "step": 52260 }, { "epoch": 1.0445226010151474, "grad_norm": 0.14760886132717133, "learning_rate": 8.049078924831886e-05, "loss": 0.0278, "step": 52270 }, { "epoch": 1.0447224331561489, "grad_norm": 0.10644547641277313, "learning_rate": 8.048259830809517e-05, "loss": 0.0312, "step": 52280 }, { "epoch": 1.0449222652971504, "grad_norm": 0.17751368880271912, "learning_rate": 8.04744060657109e-05, "loss": 0.0416, "step": 52290 }, { "epoch": 1.0451220974381519, "grad_norm": 0.12775802612304688, "learning_rate": 8.046621252151595e-05, "loss": 0.0363, "step": 52300 }, { "epoch": 1.0453219295791536, "grad_norm": 0.11969784647226334, "learning_rate": 8.04580176758604e-05, "loss": 0.0386, "step": 52310 }, { "epoch": 1.045521761720155, "grad_norm": 0.14206549525260925, "learning_rate": 8.044982152909428e-05, "loss": 0.0367, "step": 52320 }, { "epoch": 1.0457215938611566, "grad_norm": 0.13233128190040588, "learning_rate": 8.044162408156773e-05, "loss": 0.0298, "step": 52330 }, { "epoch": 1.045921426002158, "grad_norm": 0.14804179966449738, "learning_rate": 8.04334253336309e-05, "loss": 0.0376, "step": 52340 }, { "epoch": 1.0461212581431598, "grad_norm": 0.2070828527212143, "learning_rate": 8.042522528563406e-05, "loss": 0.0451, "step": 52350 }, { "epoch": 1.0463210902841613, "grad_norm": 0.1365921050310135, "learning_rate": 8.04170239379275e-05, "loss": 0.0358, "step": 52360 }, { "epoch": 1.0465209224251628, "grad_norm": 0.11183222383260727, "learning_rate": 8.040882129086154e-05, "loss": 0.0286, "step": 52370 }, { "epoch": 1.0467207545661643, "grad_norm": 0.14218421280384064, "learning_rate": 8.04006173447866e-05, "loss": 0.0344, "step": 52380 }, { "epoch": 1.046920586707166, "grad_norm": 0.12486056983470917, "learning_rate": 8.039241210005312e-05, "loss": 0.0341, "step": 52390 }, { "epoch": 1.0471204188481675, "grad_norm": 0.12980306148529053, "learning_rate": 8.038420555701166e-05, "loss": 0.0438, "step": 52400 }, { "epoch": 1.047320250989169, "grad_norm": 0.14043305814266205, "learning_rate": 8.037599771601274e-05, "loss": 0.0332, "step": 52410 }, { "epoch": 1.0475200831301708, "grad_norm": 0.1409132033586502, "learning_rate": 8.036778857740699e-05, "loss": 0.0398, "step": 52420 }, { "epoch": 1.0477199152711723, "grad_norm": 0.14320126175880432, "learning_rate": 8.035957814154512e-05, "loss": 0.0326, "step": 52430 }, { "epoch": 1.0479197474121738, "grad_norm": 0.20055678486824036, "learning_rate": 8.035136640877783e-05, "loss": 0.0384, "step": 52440 }, { "epoch": 1.0481195795531753, "grad_norm": 0.14236967265605927, "learning_rate": 8.034315337945593e-05, "loss": 0.037, "step": 52450 }, { "epoch": 1.048319411694177, "grad_norm": 0.1217576116323471, "learning_rate": 8.033493905393026e-05, "loss": 0.0341, "step": 52460 }, { "epoch": 1.0485192438351785, "grad_norm": 0.10849776864051819, "learning_rate": 8.032672343255171e-05, "loss": 0.0304, "step": 52470 }, { "epoch": 1.04871907597618, "grad_norm": 0.12829764187335968, "learning_rate": 8.031850651567126e-05, "loss": 0.0332, "step": 52480 }, { "epoch": 1.0489189081171815, "grad_norm": 0.1378500610589981, "learning_rate": 8.031028830363991e-05, "loss": 0.0301, "step": 52490 }, { "epoch": 1.0491187402581832, "grad_norm": 0.14310243725776672, "learning_rate": 8.030206879680871e-05, "loss": 0.0385, "step": 52500 }, { "epoch": 1.0493185723991847, "grad_norm": 0.17357587814331055, "learning_rate": 8.029384799552882e-05, "loss": 0.0381, "step": 52510 }, { "epoch": 1.0495184045401862, "grad_norm": 0.138727605342865, "learning_rate": 8.028562590015138e-05, "loss": 0.0329, "step": 52520 }, { "epoch": 1.049718236681188, "grad_norm": 0.14696675539016724, "learning_rate": 8.027740251102762e-05, "loss": 0.0349, "step": 52530 }, { "epoch": 1.0499180688221894, "grad_norm": 0.12892043590545654, "learning_rate": 8.026917782850888e-05, "loss": 0.0313, "step": 52540 }, { "epoch": 1.050117900963191, "grad_norm": 0.1779673993587494, "learning_rate": 8.026095185294646e-05, "loss": 0.03, "step": 52550 }, { "epoch": 1.0503177331041924, "grad_norm": 0.15769681334495544, "learning_rate": 8.025272458469176e-05, "loss": 0.0306, "step": 52560 }, { "epoch": 1.0505175652451941, "grad_norm": 0.15286892652511597, "learning_rate": 8.024449602409624e-05, "loss": 0.0373, "step": 52570 }, { "epoch": 1.0507173973861956, "grad_norm": 0.13888657093048096, "learning_rate": 8.023626617151141e-05, "loss": 0.0352, "step": 52580 }, { "epoch": 1.0509172295271971, "grad_norm": 0.1679425835609436, "learning_rate": 8.022803502728885e-05, "loss": 0.0353, "step": 52590 }, { "epoch": 1.0511170616681986, "grad_norm": 0.16625232994556427, "learning_rate": 8.021980259178016e-05, "loss": 0.0314, "step": 52600 }, { "epoch": 1.0513168938092003, "grad_norm": 0.14498288929462433, "learning_rate": 8.0211568865337e-05, "loss": 0.0355, "step": 52610 }, { "epoch": 1.0515167259502018, "grad_norm": 0.08819156885147095, "learning_rate": 8.020333384831113e-05, "loss": 0.0349, "step": 52620 }, { "epoch": 1.0517165580912033, "grad_norm": 0.13244466483592987, "learning_rate": 8.019509754105435e-05, "loss": 0.0341, "step": 52630 }, { "epoch": 1.0519163902322048, "grad_norm": 0.10895269364118576, "learning_rate": 8.018685994391844e-05, "loss": 0.0273, "step": 52640 }, { "epoch": 1.0521162223732066, "grad_norm": 0.16710324585437775, "learning_rate": 8.017862105725534e-05, "loss": 0.0381, "step": 52650 }, { "epoch": 1.052316054514208, "grad_norm": 0.11725321412086487, "learning_rate": 8.017038088141697e-05, "loss": 0.044, "step": 52660 }, { "epoch": 1.0525158866552096, "grad_norm": 0.11362912505865097, "learning_rate": 8.016213941675537e-05, "loss": 0.0345, "step": 52670 }, { "epoch": 1.0527157187962113, "grad_norm": 0.15195044875144958, "learning_rate": 8.01538966636226e-05, "loss": 0.0323, "step": 52680 }, { "epoch": 1.0529155509372128, "grad_norm": 0.1323440968990326, "learning_rate": 8.014565262237072e-05, "loss": 0.041, "step": 52690 }, { "epoch": 1.0531153830782143, "grad_norm": 0.11846376210451126, "learning_rate": 8.013740729335197e-05, "loss": 0.0385, "step": 52700 }, { "epoch": 1.0533152152192158, "grad_norm": 0.1597503423690796, "learning_rate": 8.012916067691854e-05, "loss": 0.0316, "step": 52710 }, { "epoch": 1.0535150473602175, "grad_norm": 0.1238831877708435, "learning_rate": 8.012091277342271e-05, "loss": 0.0333, "step": 52720 }, { "epoch": 1.053714879501219, "grad_norm": 0.14141909778118134, "learning_rate": 8.011266358321681e-05, "loss": 0.0355, "step": 52730 }, { "epoch": 1.0539147116422205, "grad_norm": 0.13443422317504883, "learning_rate": 8.010441310665326e-05, "loss": 0.0414, "step": 52740 }, { "epoch": 1.054114543783222, "grad_norm": 0.12413979321718216, "learning_rate": 8.009616134408448e-05, "loss": 0.038, "step": 52750 }, { "epoch": 1.0543143759242237, "grad_norm": 0.10691317170858383, "learning_rate": 8.008790829586297e-05, "loss": 0.0392, "step": 52760 }, { "epoch": 1.0545142080652252, "grad_norm": 0.10349132865667343, "learning_rate": 8.00796539623413e-05, "loss": 0.0254, "step": 52770 }, { "epoch": 1.0547140402062267, "grad_norm": 0.1047271266579628, "learning_rate": 8.007139834387206e-05, "loss": 0.03, "step": 52780 }, { "epoch": 1.0549138723472282, "grad_norm": 0.15313078463077545, "learning_rate": 8.006314144080793e-05, "loss": 0.0324, "step": 52790 }, { "epoch": 1.05511370448823, "grad_norm": 0.14804232120513916, "learning_rate": 8.005488325350163e-05, "loss": 0.0344, "step": 52800 }, { "epoch": 1.0553135366292314, "grad_norm": 0.10682221502065659, "learning_rate": 8.004662378230593e-05, "loss": 0.0327, "step": 52810 }, { "epoch": 1.055513368770233, "grad_norm": 0.1429201066493988, "learning_rate": 8.003836302757365e-05, "loss": 0.0426, "step": 52820 }, { "epoch": 1.0557132009112347, "grad_norm": 0.1240994855761528, "learning_rate": 8.003010098965768e-05, "loss": 0.0325, "step": 52830 }, { "epoch": 1.0559130330522362, "grad_norm": 0.1321312040090561, "learning_rate": 8.002183766891099e-05, "loss": 0.03, "step": 52840 }, { "epoch": 1.0561128651932377, "grad_norm": 0.14418087899684906, "learning_rate": 8.00135730656865e-05, "loss": 0.0369, "step": 52850 }, { "epoch": 1.0563126973342392, "grad_norm": 0.16860248148441315, "learning_rate": 8.000530718033734e-05, "loss": 0.0301, "step": 52860 }, { "epoch": 1.0565125294752409, "grad_norm": 0.1675790399312973, "learning_rate": 7.999704001321659e-05, "loss": 0.0297, "step": 52870 }, { "epoch": 1.0567123616162424, "grad_norm": 0.1412792056798935, "learning_rate": 7.998877156467738e-05, "loss": 0.0408, "step": 52880 }, { "epoch": 1.0569121937572439, "grad_norm": 0.15263554453849792, "learning_rate": 7.998050183507293e-05, "loss": 0.0389, "step": 52890 }, { "epoch": 1.0571120258982454, "grad_norm": 0.12426949292421341, "learning_rate": 7.997223082475654e-05, "loss": 0.0311, "step": 52900 }, { "epoch": 1.057311858039247, "grad_norm": 0.1352107673883438, "learning_rate": 7.996395853408149e-05, "loss": 0.0369, "step": 52910 }, { "epoch": 1.0575116901802486, "grad_norm": 0.1385776251554489, "learning_rate": 7.995568496340117e-05, "loss": 0.0338, "step": 52920 }, { "epoch": 1.05771152232125, "grad_norm": 0.1282154619693756, "learning_rate": 7.994741011306903e-05, "loss": 0.0301, "step": 52930 }, { "epoch": 1.0579113544622518, "grad_norm": 0.14389421045780182, "learning_rate": 7.993913398343856e-05, "loss": 0.0294, "step": 52940 }, { "epoch": 1.0581111866032533, "grad_norm": 0.1301572620868683, "learning_rate": 7.993085657486327e-05, "loss": 0.0316, "step": 52950 }, { "epoch": 1.0583110187442548, "grad_norm": 0.12746784090995789, "learning_rate": 7.992257788769679e-05, "loss": 0.0334, "step": 52960 }, { "epoch": 1.0585108508852563, "grad_norm": 0.16313670575618744, "learning_rate": 7.991429792229272e-05, "loss": 0.0327, "step": 52970 }, { "epoch": 1.058710683026258, "grad_norm": 0.12620681524276733, "learning_rate": 7.990601667900482e-05, "loss": 0.0249, "step": 52980 }, { "epoch": 1.0589105151672595, "grad_norm": 0.12285923212766647, "learning_rate": 7.989773415818682e-05, "loss": 0.046, "step": 52990 }, { "epoch": 1.059110347308261, "grad_norm": 0.11685900390148163, "learning_rate": 7.988945036019255e-05, "loss": 0.0339, "step": 53000 }, { "epoch": 1.0593101794492625, "grad_norm": 0.1152142733335495, "learning_rate": 7.988116528537586e-05, "loss": 0.0297, "step": 53010 }, { "epoch": 1.0595100115902643, "grad_norm": 0.17034655809402466, "learning_rate": 7.987287893409068e-05, "loss": 0.0404, "step": 53020 }, { "epoch": 1.0597098437312658, "grad_norm": 0.1399058848619461, "learning_rate": 7.9864591306691e-05, "loss": 0.0419, "step": 53030 }, { "epoch": 1.0599096758722673, "grad_norm": 0.12636372447013855, "learning_rate": 7.985630240353082e-05, "loss": 0.042, "step": 53040 }, { "epoch": 1.0601095080132688, "grad_norm": 0.1503954380750656, "learning_rate": 7.984801222496427e-05, "loss": 0.0282, "step": 53050 }, { "epoch": 1.0603093401542705, "grad_norm": 0.16690590977668762, "learning_rate": 7.983972077134547e-05, "loss": 0.0292, "step": 53060 }, { "epoch": 1.060509172295272, "grad_norm": 0.15978501737117767, "learning_rate": 7.983142804302861e-05, "loss": 0.0401, "step": 53070 }, { "epoch": 1.0607090044362735, "grad_norm": 0.12895861268043518, "learning_rate": 7.982313404036795e-05, "loss": 0.0329, "step": 53080 }, { "epoch": 1.0609088365772752, "grad_norm": 0.1386743038892746, "learning_rate": 7.981483876371777e-05, "loss": 0.0324, "step": 53090 }, { "epoch": 1.0611086687182767, "grad_norm": 0.11897939443588257, "learning_rate": 7.980654221343249e-05, "loss": 0.0448, "step": 53100 }, { "epoch": 1.0613085008592782, "grad_norm": 0.14122317731380463, "learning_rate": 7.979824438986645e-05, "loss": 0.0329, "step": 53110 }, { "epoch": 1.0615083330002797, "grad_norm": 0.12972450256347656, "learning_rate": 7.978994529337418e-05, "loss": 0.0303, "step": 53120 }, { "epoch": 1.0617081651412814, "grad_norm": 0.14787134528160095, "learning_rate": 7.978164492431014e-05, "loss": 0.0351, "step": 53130 }, { "epoch": 1.061907997282283, "grad_norm": 0.15897901356220245, "learning_rate": 7.977334328302897e-05, "loss": 0.0376, "step": 53140 }, { "epoch": 1.0621078294232844, "grad_norm": 0.15359237790107727, "learning_rate": 7.976504036988524e-05, "loss": 0.0442, "step": 53150 }, { "epoch": 1.062307661564286, "grad_norm": 0.21576610207557678, "learning_rate": 7.975673618523369e-05, "loss": 0.0358, "step": 53160 }, { "epoch": 1.0625074937052876, "grad_norm": 0.11827173084020615, "learning_rate": 7.974843072942902e-05, "loss": 0.0323, "step": 53170 }, { "epoch": 1.0627073258462891, "grad_norm": 0.12380427122116089, "learning_rate": 7.974012400282604e-05, "loss": 0.0339, "step": 53180 }, { "epoch": 1.0629071579872906, "grad_norm": 0.1300976276397705, "learning_rate": 7.973181600577962e-05, "loss": 0.0309, "step": 53190 }, { "epoch": 1.0631069901282921, "grad_norm": 0.17339587211608887, "learning_rate": 7.97235067386446e-05, "loss": 0.0397, "step": 53200 }, { "epoch": 1.0633068222692938, "grad_norm": 0.12546372413635254, "learning_rate": 7.9715196201776e-05, "loss": 0.0346, "step": 53210 }, { "epoch": 1.0635066544102953, "grad_norm": 0.11825010180473328, "learning_rate": 7.97068843955288e-05, "loss": 0.0331, "step": 53220 }, { "epoch": 1.0637064865512968, "grad_norm": 0.1408223658800125, "learning_rate": 7.969857132025809e-05, "loss": 0.0306, "step": 53230 }, { "epoch": 1.0639063186922986, "grad_norm": 0.12526628375053406, "learning_rate": 7.969025697631896e-05, "loss": 0.0325, "step": 53240 }, { "epoch": 1.0641061508333, "grad_norm": 0.12492303550243378, "learning_rate": 7.968194136406657e-05, "loss": 0.0329, "step": 53250 }, { "epoch": 1.0643059829743016, "grad_norm": 0.1237444281578064, "learning_rate": 7.96736244838562e-05, "loss": 0.0431, "step": 53260 }, { "epoch": 1.064505815115303, "grad_norm": 0.12880651652812958, "learning_rate": 7.96653063360431e-05, "loss": 0.0322, "step": 53270 }, { "epoch": 1.0647056472563048, "grad_norm": 0.1969222128391266, "learning_rate": 7.96569869209826e-05, "loss": 0.0481, "step": 53280 }, { "epoch": 1.0649054793973063, "grad_norm": 0.1545824408531189, "learning_rate": 7.964866623903011e-05, "loss": 0.0378, "step": 53290 }, { "epoch": 1.0651053115383078, "grad_norm": 0.0973283052444458, "learning_rate": 7.964034429054105e-05, "loss": 0.0393, "step": 53300 }, { "epoch": 1.0653051436793093, "grad_norm": 0.14125265181064606, "learning_rate": 7.963202107587096e-05, "loss": 0.0321, "step": 53310 }, { "epoch": 1.065504975820311, "grad_norm": 0.13595882058143616, "learning_rate": 7.962369659537536e-05, "loss": 0.0333, "step": 53320 }, { "epoch": 1.0657048079613125, "grad_norm": 0.13095322251319885, "learning_rate": 7.961537084940985e-05, "loss": 0.0306, "step": 53330 }, { "epoch": 1.065904640102314, "grad_norm": 0.12008914351463318, "learning_rate": 7.960704383833011e-05, "loss": 0.0325, "step": 53340 }, { "epoch": 1.0661044722433157, "grad_norm": 0.11558952927589417, "learning_rate": 7.959871556249183e-05, "loss": 0.0255, "step": 53350 }, { "epoch": 1.0663043043843172, "grad_norm": 0.12394674122333527, "learning_rate": 7.959038602225081e-05, "loss": 0.0422, "step": 53360 }, { "epoch": 1.0665041365253187, "grad_norm": 0.11775501072406769, "learning_rate": 7.958205521796286e-05, "loss": 0.0333, "step": 53370 }, { "epoch": 1.0667039686663202, "grad_norm": 0.1308177411556244, "learning_rate": 7.957372314998384e-05, "loss": 0.029, "step": 53380 }, { "epoch": 1.066903800807322, "grad_norm": 0.11598023772239685, "learning_rate": 7.956538981866972e-05, "loss": 0.0355, "step": 53390 }, { "epoch": 1.0671036329483234, "grad_norm": 0.13344193994998932, "learning_rate": 7.955705522437645e-05, "loss": 0.0382, "step": 53400 }, { "epoch": 1.067303465089325, "grad_norm": 0.1452046036720276, "learning_rate": 7.954871936746007e-05, "loss": 0.0394, "step": 53410 }, { "epoch": 1.0675032972303264, "grad_norm": 0.11306744068861008, "learning_rate": 7.954038224827669e-05, "loss": 0.0426, "step": 53420 }, { "epoch": 1.0677031293713282, "grad_norm": 0.13399820029735565, "learning_rate": 7.953204386718245e-05, "loss": 0.0375, "step": 53430 }, { "epoch": 1.0679029615123297, "grad_norm": 0.14616219699382782, "learning_rate": 7.952370422453355e-05, "loss": 0.0406, "step": 53440 }, { "epoch": 1.0681027936533312, "grad_norm": 0.12856759130954742, "learning_rate": 7.951536332068623e-05, "loss": 0.0295, "step": 53450 }, { "epoch": 1.0683026257943327, "grad_norm": 0.13093829154968262, "learning_rate": 7.950702115599682e-05, "loss": 0.0316, "step": 53460 }, { "epoch": 1.0685024579353344, "grad_norm": 0.14506012201309204, "learning_rate": 7.949867773082168e-05, "loss": 0.0345, "step": 53470 }, { "epoch": 1.0687022900763359, "grad_norm": 0.13262438774108887, "learning_rate": 7.949033304551719e-05, "loss": 0.0301, "step": 53480 }, { "epoch": 1.0689021222173374, "grad_norm": 0.14074580371379852, "learning_rate": 7.948198710043986e-05, "loss": 0.0304, "step": 53490 }, { "epoch": 1.0691019543583389, "grad_norm": 0.10822633653879166, "learning_rate": 7.947363989594621e-05, "loss": 0.0383, "step": 53500 }, { "epoch": 1.0693017864993406, "grad_norm": 0.12315014004707336, "learning_rate": 7.946529143239281e-05, "loss": 0.037, "step": 53510 }, { "epoch": 1.069501618640342, "grad_norm": 0.12410632520914078, "learning_rate": 7.945694171013628e-05, "loss": 0.032, "step": 53520 }, { "epoch": 1.0697014507813436, "grad_norm": 0.12806181609630585, "learning_rate": 7.944859072953333e-05, "loss": 0.0322, "step": 53530 }, { "epoch": 1.0699012829223453, "grad_norm": 0.14774256944656372, "learning_rate": 7.944023849094066e-05, "loss": 0.0363, "step": 53540 }, { "epoch": 1.0701011150633468, "grad_norm": 0.13990488648414612, "learning_rate": 7.94318849947151e-05, "loss": 0.0298, "step": 53550 }, { "epoch": 1.0703009472043483, "grad_norm": 0.16475529968738556, "learning_rate": 7.942353024121347e-05, "loss": 0.0364, "step": 53560 }, { "epoch": 1.0705007793453498, "grad_norm": 0.15424907207489014, "learning_rate": 7.94151742307927e-05, "loss": 0.0426, "step": 53570 }, { "epoch": 1.0707006114863515, "grad_norm": 0.11292648315429688, "learning_rate": 7.940681696380969e-05, "loss": 0.0306, "step": 53580 }, { "epoch": 1.070900443627353, "grad_norm": 0.14923423528671265, "learning_rate": 7.939845844062153e-05, "loss": 0.0368, "step": 53590 }, { "epoch": 1.0711002757683545, "grad_norm": 0.12179412692785263, "learning_rate": 7.939009866158519e-05, "loss": 0.036, "step": 53600 }, { "epoch": 1.071300107909356, "grad_norm": 0.11850275844335556, "learning_rate": 7.938173762705785e-05, "loss": 0.0433, "step": 53610 }, { "epoch": 1.0714999400503578, "grad_norm": 0.16597826778888702, "learning_rate": 7.937337533739663e-05, "loss": 0.0307, "step": 53620 }, { "epoch": 1.0716997721913593, "grad_norm": 0.15326327085494995, "learning_rate": 7.93650117929588e-05, "loss": 0.0334, "step": 53630 }, { "epoch": 1.0718996043323608, "grad_norm": 0.13188838958740234, "learning_rate": 7.935664699410161e-05, "loss": 0.0352, "step": 53640 }, { "epoch": 1.0720994364733625, "grad_norm": 0.11265552788972855, "learning_rate": 7.934828094118239e-05, "loss": 0.0379, "step": 53650 }, { "epoch": 1.072299268614364, "grad_norm": 0.12078116089105606, "learning_rate": 7.933991363455852e-05, "loss": 0.0346, "step": 53660 }, { "epoch": 1.0724991007553655, "grad_norm": 0.16073642671108246, "learning_rate": 7.933154507458742e-05, "loss": 0.0507, "step": 53670 }, { "epoch": 1.072698932896367, "grad_norm": 0.14564120769500732, "learning_rate": 7.932317526162662e-05, "loss": 0.0334, "step": 53680 }, { "epoch": 1.0728987650373687, "grad_norm": 0.1180221363902092, "learning_rate": 7.931480419603364e-05, "loss": 0.0316, "step": 53690 }, { "epoch": 1.0730985971783702, "grad_norm": 0.25484493374824524, "learning_rate": 7.930643187816607e-05, "loss": 0.0393, "step": 53700 }, { "epoch": 1.0732984293193717, "grad_norm": 0.12024933844804764, "learning_rate": 7.929805830838157e-05, "loss": 0.0342, "step": 53710 }, { "epoch": 1.0734982614603732, "grad_norm": 0.1445760726928711, "learning_rate": 7.928968348703785e-05, "loss": 0.0467, "step": 53720 }, { "epoch": 1.073698093601375, "grad_norm": 0.11646277457475662, "learning_rate": 7.928130741449264e-05, "loss": 0.034, "step": 53730 }, { "epoch": 1.0738979257423764, "grad_norm": 0.12557227909564972, "learning_rate": 7.92729300911038e-05, "loss": 0.0379, "step": 53740 }, { "epoch": 1.074097757883378, "grad_norm": 0.11659248918294907, "learning_rate": 7.926455151722914e-05, "loss": 0.0418, "step": 53750 }, { "epoch": 1.0742975900243796, "grad_norm": 0.13887818157672882, "learning_rate": 7.925617169322659e-05, "loss": 0.0372, "step": 53760 }, { "epoch": 1.0744974221653811, "grad_norm": 0.1747179925441742, "learning_rate": 7.924779061945415e-05, "loss": 0.0301, "step": 53770 }, { "epoch": 1.0746972543063826, "grad_norm": 0.1255679428577423, "learning_rate": 7.923940829626982e-05, "loss": 0.0299, "step": 53780 }, { "epoch": 1.0748970864473841, "grad_norm": 0.1706090271472931, "learning_rate": 7.923102472403167e-05, "loss": 0.0311, "step": 53790 }, { "epoch": 1.0750969185883859, "grad_norm": 0.1314888447523117, "learning_rate": 7.922263990309786e-05, "loss": 0.0382, "step": 53800 }, { "epoch": 1.0752967507293874, "grad_norm": 0.13291078805923462, "learning_rate": 7.921425383382655e-05, "loss": 0.0357, "step": 53810 }, { "epoch": 1.0754965828703889, "grad_norm": 0.11737990379333496, "learning_rate": 7.920586651657598e-05, "loss": 0.0355, "step": 53820 }, { "epoch": 1.0756964150113904, "grad_norm": 0.16384324431419373, "learning_rate": 7.919747795170444e-05, "loss": 0.0298, "step": 53830 }, { "epoch": 1.075896247152392, "grad_norm": 0.17919771373271942, "learning_rate": 7.91890881395703e-05, "loss": 0.0313, "step": 53840 }, { "epoch": 1.0760960792933936, "grad_norm": 0.11172698438167572, "learning_rate": 7.918069708053191e-05, "loss": 0.0301, "step": 53850 }, { "epoch": 1.076295911434395, "grad_norm": 0.13967876136302948, "learning_rate": 7.917230477494777e-05, "loss": 0.0313, "step": 53860 }, { "epoch": 1.0764957435753966, "grad_norm": 0.14982926845550537, "learning_rate": 7.916391122317636e-05, "loss": 0.0373, "step": 53870 }, { "epoch": 1.0766955757163983, "grad_norm": 0.12782084941864014, "learning_rate": 7.915551642557624e-05, "loss": 0.046, "step": 53880 }, { "epoch": 1.0768954078573998, "grad_norm": 0.13184984028339386, "learning_rate": 7.9147120382506e-05, "loss": 0.0448, "step": 53890 }, { "epoch": 1.0770952399984013, "grad_norm": 0.11580993980169296, "learning_rate": 7.913872309432433e-05, "loss": 0.0258, "step": 53900 }, { "epoch": 1.0772950721394028, "grad_norm": 0.14701174199581146, "learning_rate": 7.913032456138994e-05, "loss": 0.0387, "step": 53910 }, { "epoch": 1.0774949042804045, "grad_norm": 0.12088993191719055, "learning_rate": 7.91219247840616e-05, "loss": 0.0358, "step": 53920 }, { "epoch": 1.077694736421406, "grad_norm": 0.12029127776622772, "learning_rate": 7.911352376269814e-05, "loss": 0.0363, "step": 53930 }, { "epoch": 1.0778945685624075, "grad_norm": 0.1541687697172165, "learning_rate": 7.910512149765842e-05, "loss": 0.0319, "step": 53940 }, { "epoch": 1.0780944007034092, "grad_norm": 0.11103325337171555, "learning_rate": 7.909671798930138e-05, "loss": 0.0322, "step": 53950 }, { "epoch": 1.0782942328444107, "grad_norm": 0.12757699191570282, "learning_rate": 7.9088313237986e-05, "loss": 0.0319, "step": 53960 }, { "epoch": 1.0784940649854122, "grad_norm": 0.13990812003612518, "learning_rate": 7.907990724407131e-05, "loss": 0.0381, "step": 53970 }, { "epoch": 1.0786938971264137, "grad_norm": 0.13661707937717438, "learning_rate": 7.907150000791641e-05, "loss": 0.0349, "step": 53980 }, { "epoch": 1.0788937292674154, "grad_norm": 0.16046665608882904, "learning_rate": 7.906309152988043e-05, "loss": 0.0379, "step": 53990 }, { "epoch": 1.079093561408417, "grad_norm": 0.14059051871299744, "learning_rate": 7.905468181032259e-05, "loss": 0.0339, "step": 54000 }, { "epoch": 1.0792933935494184, "grad_norm": 0.18926236033439636, "learning_rate": 7.90462708496021e-05, "loss": 0.041, "step": 54010 }, { "epoch": 1.07949322569042, "grad_norm": 0.12738436460494995, "learning_rate": 7.903785864807827e-05, "loss": 0.0396, "step": 54020 }, { "epoch": 1.0796930578314217, "grad_norm": 0.16286152601242065, "learning_rate": 7.902944520611048e-05, "loss": 0.0347, "step": 54030 }, { "epoch": 1.0798928899724232, "grad_norm": 0.09587746113538742, "learning_rate": 7.902103052405811e-05, "loss": 0.0363, "step": 54040 }, { "epoch": 1.0800927221134247, "grad_norm": 0.14554153382778168, "learning_rate": 7.901261460228062e-05, "loss": 0.0325, "step": 54050 }, { "epoch": 1.0802925542544264, "grad_norm": 0.23957014083862305, "learning_rate": 7.900419744113757e-05, "loss": 0.0323, "step": 54060 }, { "epoch": 1.0804923863954279, "grad_norm": 0.2380477786064148, "learning_rate": 7.899577904098846e-05, "loss": 0.0287, "step": 54070 }, { "epoch": 1.0806922185364294, "grad_norm": 0.11456810683012009, "learning_rate": 7.898735940219292e-05, "loss": 0.035, "step": 54080 }, { "epoch": 1.0808920506774309, "grad_norm": 0.1280011534690857, "learning_rate": 7.897893852511067e-05, "loss": 0.0314, "step": 54090 }, { "epoch": 1.0810918828184326, "grad_norm": 0.1755528450012207, "learning_rate": 7.897051641010138e-05, "loss": 0.0367, "step": 54100 }, { "epoch": 1.081291714959434, "grad_norm": 0.13534076511859894, "learning_rate": 7.896209305752487e-05, "loss": 0.0342, "step": 54110 }, { "epoch": 1.0814915471004356, "grad_norm": 0.1330684870481491, "learning_rate": 7.895366846774093e-05, "loss": 0.0302, "step": 54120 }, { "epoch": 1.081691379241437, "grad_norm": 0.13583895564079285, "learning_rate": 7.894524264110947e-05, "loss": 0.0403, "step": 54130 }, { "epoch": 1.0818912113824388, "grad_norm": 0.11988236755132675, "learning_rate": 7.893681557799041e-05, "loss": 0.0321, "step": 54140 }, { "epoch": 1.0820910435234403, "grad_norm": 0.14521487057209015, "learning_rate": 7.892838727874377e-05, "loss": 0.0334, "step": 54150 }, { "epoch": 1.0822908756644418, "grad_norm": 0.15230335295200348, "learning_rate": 7.891995774372956e-05, "loss": 0.0355, "step": 54160 }, { "epoch": 1.0824907078054435, "grad_norm": 0.13903279602527618, "learning_rate": 7.891152697330788e-05, "loss": 0.0321, "step": 54170 }, { "epoch": 1.082690539946445, "grad_norm": 0.13560353219509125, "learning_rate": 7.890309496783888e-05, "loss": 0.0415, "step": 54180 }, { "epoch": 1.0828903720874465, "grad_norm": 0.13963541388511658, "learning_rate": 7.889466172768273e-05, "loss": 0.0381, "step": 54190 }, { "epoch": 1.083090204228448, "grad_norm": 0.11859899014234543, "learning_rate": 7.888622725319977e-05, "loss": 0.0329, "step": 54200 }, { "epoch": 1.0832900363694498, "grad_norm": 0.09966887533664703, "learning_rate": 7.88777915447502e-05, "loss": 0.042, "step": 54210 }, { "epoch": 1.0834898685104513, "grad_norm": 0.14942489564418793, "learning_rate": 7.886935460269445e-05, "loss": 0.0349, "step": 54220 }, { "epoch": 1.0836897006514528, "grad_norm": 0.13556969165802002, "learning_rate": 7.886091642739289e-05, "loss": 0.0327, "step": 54230 }, { "epoch": 1.0838895327924543, "grad_norm": 0.15758492052555084, "learning_rate": 7.885247701920602e-05, "loss": 0.0359, "step": 54240 }, { "epoch": 1.084089364933456, "grad_norm": 0.11251366883516312, "learning_rate": 7.884403637849431e-05, "loss": 0.0349, "step": 54250 }, { "epoch": 1.0842891970744575, "grad_norm": 0.13727548718452454, "learning_rate": 7.883559450561837e-05, "loss": 0.0408, "step": 54260 }, { "epoch": 1.084489029215459, "grad_norm": 0.14286354184150696, "learning_rate": 7.882715140093882e-05, "loss": 0.0351, "step": 54270 }, { "epoch": 1.0846888613564605, "grad_norm": 0.13645896315574646, "learning_rate": 7.881870706481629e-05, "loss": 0.0379, "step": 54280 }, { "epoch": 1.0848886934974622, "grad_norm": 0.1583438366651535, "learning_rate": 7.881026149761155e-05, "loss": 0.0437, "step": 54290 }, { "epoch": 1.0850885256384637, "grad_norm": 0.1370357722043991, "learning_rate": 7.880181469968537e-05, "loss": 0.0344, "step": 54300 }, { "epoch": 1.0852883577794652, "grad_norm": 0.1268109530210495, "learning_rate": 7.879336667139856e-05, "loss": 0.0386, "step": 54310 }, { "epoch": 1.0854881899204667, "grad_norm": 0.11424435675144196, "learning_rate": 7.878491741311203e-05, "loss": 0.0296, "step": 54320 }, { "epoch": 1.0856880220614684, "grad_norm": 0.12789887189865112, "learning_rate": 7.877646692518673e-05, "loss": 0.0291, "step": 54330 }, { "epoch": 1.08588785420247, "grad_norm": 0.12265560030937195, "learning_rate": 7.876801520798358e-05, "loss": 0.0355, "step": 54340 }, { "epoch": 1.0860876863434714, "grad_norm": 0.1573442965745926, "learning_rate": 7.875956226186369e-05, "loss": 0.0455, "step": 54350 }, { "epoch": 1.0862875184844731, "grad_norm": 0.14004097878932953, "learning_rate": 7.875110808718816e-05, "loss": 0.0411, "step": 54360 }, { "epoch": 1.0864873506254746, "grad_norm": 0.120174340903759, "learning_rate": 7.874265268431808e-05, "loss": 0.0351, "step": 54370 }, { "epoch": 1.0866871827664761, "grad_norm": 0.14403027296066284, "learning_rate": 7.873419605361469e-05, "loss": 0.0349, "step": 54380 }, { "epoch": 1.0868870149074776, "grad_norm": 0.15454447269439697, "learning_rate": 7.872573819543923e-05, "loss": 0.0387, "step": 54390 }, { "epoch": 1.0870868470484794, "grad_norm": 0.13332095742225647, "learning_rate": 7.871727911015298e-05, "loss": 0.0304, "step": 54400 }, { "epoch": 1.0872866791894809, "grad_norm": 0.16506490111351013, "learning_rate": 7.870881879811735e-05, "loss": 0.0267, "step": 54410 }, { "epoch": 1.0874865113304824, "grad_norm": 0.14180469512939453, "learning_rate": 7.870035725969371e-05, "loss": 0.0337, "step": 54420 }, { "epoch": 1.0876863434714839, "grad_norm": 0.15241935849189758, "learning_rate": 7.869189449524355e-05, "loss": 0.0359, "step": 54430 }, { "epoch": 1.0878861756124856, "grad_norm": 0.10223176330327988, "learning_rate": 7.868343050512835e-05, "loss": 0.0311, "step": 54440 }, { "epoch": 1.088086007753487, "grad_norm": 0.10557322949171066, "learning_rate": 7.86749652897097e-05, "loss": 0.0279, "step": 54450 }, { "epoch": 1.0882858398944886, "grad_norm": 0.145061194896698, "learning_rate": 7.866649884934921e-05, "loss": 0.0343, "step": 54460 }, { "epoch": 1.0884856720354903, "grad_norm": 0.14743949472904205, "learning_rate": 7.865803118440855e-05, "loss": 0.0344, "step": 54470 }, { "epoch": 1.0886855041764918, "grad_norm": 0.16138888895511627, "learning_rate": 7.864956229524942e-05, "loss": 0.0361, "step": 54480 }, { "epoch": 1.0888853363174933, "grad_norm": 0.14870376884937286, "learning_rate": 7.864109218223366e-05, "loss": 0.0337, "step": 54490 }, { "epoch": 1.0890851684584948, "grad_norm": 0.12475933134555817, "learning_rate": 7.863262084572303e-05, "loss": 0.0307, "step": 54500 }, { "epoch": 1.0892850005994965, "grad_norm": 0.13083094358444214, "learning_rate": 7.862414828607944e-05, "loss": 0.0319, "step": 54510 }, { "epoch": 1.089484832740498, "grad_norm": 0.14571250975131989, "learning_rate": 7.861567450366483e-05, "loss": 0.0371, "step": 54520 }, { "epoch": 1.0896846648814995, "grad_norm": 0.1254063844680786, "learning_rate": 7.860719949884117e-05, "loss": 0.0241, "step": 54530 }, { "epoch": 1.089884497022501, "grad_norm": 0.19747774302959442, "learning_rate": 7.859872327197048e-05, "loss": 0.0399, "step": 54540 }, { "epoch": 1.0900843291635027, "grad_norm": 0.1585264503955841, "learning_rate": 7.859024582341489e-05, "loss": 0.0362, "step": 54550 }, { "epoch": 1.0902841613045042, "grad_norm": 0.14557649195194244, "learning_rate": 7.858176715353654e-05, "loss": 0.036, "step": 54560 }, { "epoch": 1.0904839934455057, "grad_norm": 0.1538800746202469, "learning_rate": 7.857328726269758e-05, "loss": 0.0384, "step": 54570 }, { "epoch": 1.0906838255865074, "grad_norm": 0.13814792037010193, "learning_rate": 7.856480615126028e-05, "loss": 0.0338, "step": 54580 }, { "epoch": 1.090883657727509, "grad_norm": 0.1226753294467926, "learning_rate": 7.855632381958693e-05, "loss": 0.0376, "step": 54590 }, { "epoch": 1.0910834898685104, "grad_norm": 0.18980590999126434, "learning_rate": 7.85478402680399e-05, "loss": 0.0371, "step": 54600 }, { "epoch": 1.091283322009512, "grad_norm": 0.12808339297771454, "learning_rate": 7.853935549698157e-05, "loss": 0.0288, "step": 54610 }, { "epoch": 1.0914831541505137, "grad_norm": 0.13078077137470245, "learning_rate": 7.853086950677441e-05, "loss": 0.0319, "step": 54620 }, { "epoch": 1.0916829862915152, "grad_norm": 0.14472021162509918, "learning_rate": 7.852238229778091e-05, "loss": 0.0309, "step": 54630 }, { "epoch": 1.0918828184325167, "grad_norm": 0.14420533180236816, "learning_rate": 7.851389387036363e-05, "loss": 0.0352, "step": 54640 }, { "epoch": 1.0920826505735182, "grad_norm": 0.12235303223133087, "learning_rate": 7.85054042248852e-05, "loss": 0.041, "step": 54650 }, { "epoch": 1.0922824827145199, "grad_norm": 0.126450315117836, "learning_rate": 7.849691336170826e-05, "loss": 0.0281, "step": 54660 }, { "epoch": 1.0924823148555214, "grad_norm": 0.13962343335151672, "learning_rate": 7.848842128119554e-05, "loss": 0.0339, "step": 54670 }, { "epoch": 1.0926821469965229, "grad_norm": 0.11993565410375595, "learning_rate": 7.847992798370978e-05, "loss": 0.037, "step": 54680 }, { "epoch": 1.0928819791375244, "grad_norm": 0.1299840360879898, "learning_rate": 7.847143346961383e-05, "loss": 0.0408, "step": 54690 }, { "epoch": 1.093081811278526, "grad_norm": 0.10959307849407196, "learning_rate": 7.846293773927053e-05, "loss": 0.0394, "step": 54700 }, { "epoch": 1.0932816434195276, "grad_norm": 0.14451204240322113, "learning_rate": 7.845444079304283e-05, "loss": 0.0413, "step": 54710 }, { "epoch": 1.093481475560529, "grad_norm": 0.1223522499203682, "learning_rate": 7.844594263129368e-05, "loss": 0.0371, "step": 54720 }, { "epoch": 1.0936813077015306, "grad_norm": 0.13450919091701508, "learning_rate": 7.843744325438614e-05, "loss": 0.0421, "step": 54730 }, { "epoch": 1.0938811398425323, "grad_norm": 0.10396059602499008, "learning_rate": 7.842894266268325e-05, "loss": 0.0379, "step": 54740 }, { "epoch": 1.0940809719835338, "grad_norm": 0.12166937440633774, "learning_rate": 7.842044085654816e-05, "loss": 0.0329, "step": 54750 }, { "epoch": 1.0942808041245353, "grad_norm": 0.13628782331943512, "learning_rate": 7.841193783634406e-05, "loss": 0.0327, "step": 54760 }, { "epoch": 1.094480636265537, "grad_norm": 0.18823574483394623, "learning_rate": 7.840343360243414e-05, "loss": 0.0388, "step": 54770 }, { "epoch": 1.0946804684065385, "grad_norm": 0.12311571836471558, "learning_rate": 7.839492815518175e-05, "loss": 0.0362, "step": 54780 }, { "epoch": 1.09488030054754, "grad_norm": 0.15041911602020264, "learning_rate": 7.838642149495015e-05, "loss": 0.0399, "step": 54790 }, { "epoch": 1.0950801326885415, "grad_norm": 0.10449729114770889, "learning_rate": 7.83779136221028e-05, "loss": 0.0389, "step": 54800 }, { "epoch": 1.0952799648295433, "grad_norm": 0.14584994316101074, "learning_rate": 7.836940453700309e-05, "loss": 0.0384, "step": 54810 }, { "epoch": 1.0954797969705448, "grad_norm": 0.11706207692623138, "learning_rate": 7.836089424001456e-05, "loss": 0.0325, "step": 54820 }, { "epoch": 1.0956796291115463, "grad_norm": 0.15340253710746765, "learning_rate": 7.835238273150073e-05, "loss": 0.0351, "step": 54830 }, { "epoch": 1.0958794612525478, "grad_norm": 0.17479820549488068, "learning_rate": 7.834387001182519e-05, "loss": 0.0378, "step": 54840 }, { "epoch": 1.0960792933935495, "grad_norm": 0.16997909545898438, "learning_rate": 7.833535608135158e-05, "loss": 0.0303, "step": 54850 }, { "epoch": 1.096279125534551, "grad_norm": 0.11027609556913376, "learning_rate": 7.832684094044362e-05, "loss": 0.0328, "step": 54860 }, { "epoch": 1.0964789576755525, "grad_norm": 0.13601306080818176, "learning_rate": 7.831832458946504e-05, "loss": 0.033, "step": 54870 }, { "epoch": 1.0966787898165542, "grad_norm": 0.11025257408618927, "learning_rate": 7.830980702877969e-05, "loss": 0.0259, "step": 54880 }, { "epoch": 1.0968786219575557, "grad_norm": 0.13591082394123077, "learning_rate": 7.830128825875136e-05, "loss": 0.0376, "step": 54890 }, { "epoch": 1.0970784540985572, "grad_norm": 0.12916679680347443, "learning_rate": 7.829276827974398e-05, "loss": 0.0341, "step": 54900 }, { "epoch": 1.0972782862395587, "grad_norm": 0.12966948747634888, "learning_rate": 7.828424709212154e-05, "loss": 0.0341, "step": 54910 }, { "epoch": 1.0974781183805604, "grad_norm": 0.147922545671463, "learning_rate": 7.827572469624802e-05, "loss": 0.0461, "step": 54920 }, { "epoch": 1.097677950521562, "grad_norm": 0.15406911075115204, "learning_rate": 7.826720109248747e-05, "loss": 0.0377, "step": 54930 }, { "epoch": 1.0978777826625634, "grad_norm": 0.15872041881084442, "learning_rate": 7.825867628120402e-05, "loss": 0.032, "step": 54940 }, { "epoch": 1.098077614803565, "grad_norm": 0.12070485949516296, "learning_rate": 7.825015026276185e-05, "loss": 0.0305, "step": 54950 }, { "epoch": 1.0982774469445666, "grad_norm": 0.13864901661872864, "learning_rate": 7.824162303752513e-05, "loss": 0.0323, "step": 54960 }, { "epoch": 1.0984772790855681, "grad_norm": 0.12424442917108536, "learning_rate": 7.823309460585818e-05, "loss": 0.0382, "step": 54970 }, { "epoch": 1.0986771112265696, "grad_norm": 0.15436765551567078, "learning_rate": 7.822456496812529e-05, "loss": 0.0397, "step": 54980 }, { "epoch": 1.0988769433675714, "grad_norm": 0.21681424975395203, "learning_rate": 7.821603412469082e-05, "loss": 0.0421, "step": 54990 }, { "epoch": 1.0990767755085729, "grad_norm": 0.12493313103914261, "learning_rate": 7.820750207591923e-05, "loss": 0.0373, "step": 55000 }, { "epoch": 1.0992766076495744, "grad_norm": 0.15986132621765137, "learning_rate": 7.819896882217496e-05, "loss": 0.033, "step": 55010 }, { "epoch": 1.0994764397905759, "grad_norm": 0.13903480768203735, "learning_rate": 7.819043436382253e-05, "loss": 0.0369, "step": 55020 }, { "epoch": 1.0996762719315776, "grad_norm": 0.12971940636634827, "learning_rate": 7.818189870122656e-05, "loss": 0.0376, "step": 55030 }, { "epoch": 1.099876104072579, "grad_norm": 0.17501656711101532, "learning_rate": 7.817336183475163e-05, "loss": 0.0386, "step": 55040 }, { "epoch": 1.1000759362135806, "grad_norm": 0.13596069812774658, "learning_rate": 7.816482376476244e-05, "loss": 0.0268, "step": 55050 }, { "epoch": 1.100275768354582, "grad_norm": 0.140250563621521, "learning_rate": 7.815628449162371e-05, "loss": 0.0306, "step": 55060 }, { "epoch": 1.1004756004955838, "grad_norm": 0.12094253301620483, "learning_rate": 7.814774401570025e-05, "loss": 0.0361, "step": 55070 }, { "epoch": 1.1006754326365853, "grad_norm": 0.14649684727191925, "learning_rate": 7.813920233735688e-05, "loss": 0.0319, "step": 55080 }, { "epoch": 1.1008752647775868, "grad_norm": 0.11590681970119476, "learning_rate": 7.813065945695845e-05, "loss": 0.0238, "step": 55090 }, { "epoch": 1.1010750969185883, "grad_norm": 0.15989676117897034, "learning_rate": 7.812211537486994e-05, "loss": 0.0355, "step": 55100 }, { "epoch": 1.10127492905959, "grad_norm": 0.13799890875816345, "learning_rate": 7.811357009145633e-05, "loss": 0.0341, "step": 55110 }, { "epoch": 1.1014747612005915, "grad_norm": 0.1506059169769287, "learning_rate": 7.810502360708265e-05, "loss": 0.0347, "step": 55120 }, { "epoch": 1.101674593341593, "grad_norm": 0.13477317988872528, "learning_rate": 7.8096475922114e-05, "loss": 0.0352, "step": 55130 }, { "epoch": 1.1018744254825945, "grad_norm": 0.1898431032896042, "learning_rate": 7.80879270369155e-05, "loss": 0.0352, "step": 55140 }, { "epoch": 1.1020742576235962, "grad_norm": 0.11511868983507156, "learning_rate": 7.807937695185235e-05, "loss": 0.0299, "step": 55150 }, { "epoch": 1.1022740897645977, "grad_norm": 0.1843544989824295, "learning_rate": 7.807082566728982e-05, "loss": 0.0312, "step": 55160 }, { "epoch": 1.1024739219055992, "grad_norm": 0.13511870801448822, "learning_rate": 7.806227318359316e-05, "loss": 0.0312, "step": 55170 }, { "epoch": 1.102673754046601, "grad_norm": 0.10464093834161758, "learning_rate": 7.805371950112776e-05, "loss": 0.0299, "step": 55180 }, { "epoch": 1.1028735861876025, "grad_norm": 0.14613690972328186, "learning_rate": 7.8045164620259e-05, "loss": 0.0377, "step": 55190 }, { "epoch": 1.103073418328604, "grad_norm": 0.14844565093517303, "learning_rate": 7.803660854135233e-05, "loss": 0.0349, "step": 55200 }, { "epoch": 1.1032732504696054, "grad_norm": 0.15594209730625153, "learning_rate": 7.802805126477323e-05, "loss": 0.0359, "step": 55210 }, { "epoch": 1.1034730826106072, "grad_norm": 0.13245506584644318, "learning_rate": 7.801949279088728e-05, "loss": 0.0348, "step": 55220 }, { "epoch": 1.1036729147516087, "grad_norm": 0.14548195898532867, "learning_rate": 7.801093312006007e-05, "loss": 0.0349, "step": 55230 }, { "epoch": 1.1038727468926102, "grad_norm": 0.11978600919246674, "learning_rate": 7.800237225265726e-05, "loss": 0.0269, "step": 55240 }, { "epoch": 1.1040725790336117, "grad_norm": 0.13336126506328583, "learning_rate": 7.799381018904455e-05, "loss": 0.035, "step": 55250 }, { "epoch": 1.1042724111746134, "grad_norm": 0.12690576910972595, "learning_rate": 7.798524692958769e-05, "loss": 0.0349, "step": 55260 }, { "epoch": 1.1044722433156149, "grad_norm": 0.11261066049337387, "learning_rate": 7.79766824746525e-05, "loss": 0.036, "step": 55270 }, { "epoch": 1.1046720754566164, "grad_norm": 0.105218805372715, "learning_rate": 7.796811682460481e-05, "loss": 0.0324, "step": 55280 }, { "epoch": 1.104871907597618, "grad_norm": 0.12278818339109421, "learning_rate": 7.795954997981056e-05, "loss": 0.0397, "step": 55290 }, { "epoch": 1.1050717397386196, "grad_norm": 0.15203982591629028, "learning_rate": 7.79509819406357e-05, "loss": 0.042, "step": 55300 }, { "epoch": 1.105271571879621, "grad_norm": 0.1186981350183487, "learning_rate": 7.794241270744624e-05, "loss": 0.034, "step": 55310 }, { "epoch": 1.1054714040206226, "grad_norm": 0.13112425804138184, "learning_rate": 7.793384228060824e-05, "loss": 0.0354, "step": 55320 }, { "epoch": 1.1056712361616243, "grad_norm": 0.176078200340271, "learning_rate": 7.792527066048782e-05, "loss": 0.0363, "step": 55330 }, { "epoch": 1.1058710683026258, "grad_norm": 0.11814223974943161, "learning_rate": 7.791669784745112e-05, "loss": 0.0421, "step": 55340 }, { "epoch": 1.1060709004436273, "grad_norm": 0.1279630959033966, "learning_rate": 7.790812384186439e-05, "loss": 0.0369, "step": 55350 }, { "epoch": 1.1062707325846288, "grad_norm": 0.16834571957588196, "learning_rate": 7.789954864409388e-05, "loss": 0.0321, "step": 55360 }, { "epoch": 1.1064705647256305, "grad_norm": 0.11748398095369339, "learning_rate": 7.789097225450588e-05, "loss": 0.0281, "step": 55370 }, { "epoch": 1.106670396866632, "grad_norm": 0.15046831965446472, "learning_rate": 7.788239467346681e-05, "loss": 0.0333, "step": 55380 }, { "epoch": 1.1068702290076335, "grad_norm": 0.1147216260433197, "learning_rate": 7.787381590134306e-05, "loss": 0.0382, "step": 55390 }, { "epoch": 1.107070061148635, "grad_norm": 0.14478252828121185, "learning_rate": 7.786523593850108e-05, "loss": 0.029, "step": 55400 }, { "epoch": 1.1072698932896368, "grad_norm": 0.14066870510578156, "learning_rate": 7.785665478530743e-05, "loss": 0.0281, "step": 55410 }, { "epoch": 1.1074697254306383, "grad_norm": 0.11052046716213226, "learning_rate": 7.784807244212863e-05, "loss": 0.0314, "step": 55420 }, { "epoch": 1.1076695575716398, "grad_norm": 0.12448210269212723, "learning_rate": 7.783948890933137e-05, "loss": 0.035, "step": 55430 }, { "epoch": 1.1078693897126415, "grad_norm": 0.12643057107925415, "learning_rate": 7.783090418728228e-05, "loss": 0.0279, "step": 55440 }, { "epoch": 1.108069221853643, "grad_norm": 0.1420607715845108, "learning_rate": 7.782231827634808e-05, "loss": 0.0333, "step": 55450 }, { "epoch": 1.1082690539946445, "grad_norm": 0.2140382081270218, "learning_rate": 7.781373117689555e-05, "loss": 0.0383, "step": 55460 }, { "epoch": 1.108468886135646, "grad_norm": 0.13623103499412537, "learning_rate": 7.780514288929154e-05, "loss": 0.0382, "step": 55470 }, { "epoch": 1.1086687182766477, "grad_norm": 0.11106019467115402, "learning_rate": 7.779655341390289e-05, "loss": 0.0354, "step": 55480 }, { "epoch": 1.1088685504176492, "grad_norm": 0.1469677984714508, "learning_rate": 7.778796275109654e-05, "loss": 0.0365, "step": 55490 }, { "epoch": 1.1090683825586507, "grad_norm": 0.13654470443725586, "learning_rate": 7.777937090123948e-05, "loss": 0.0349, "step": 55500 }, { "epoch": 1.1092682146996522, "grad_norm": 0.10960444062948227, "learning_rate": 7.777077786469873e-05, "loss": 0.0364, "step": 55510 }, { "epoch": 1.109468046840654, "grad_norm": 0.13120219111442566, "learning_rate": 7.776218364184136e-05, "loss": 0.0345, "step": 55520 }, { "epoch": 1.1096678789816554, "grad_norm": 0.13112515211105347, "learning_rate": 7.775358823303449e-05, "loss": 0.0354, "step": 55530 }, { "epoch": 1.109867711122657, "grad_norm": 0.15453827381134033, "learning_rate": 7.774499163864535e-05, "loss": 0.0356, "step": 55540 }, { "epoch": 1.1100675432636584, "grad_norm": 0.12919536232948303, "learning_rate": 7.77363938590411e-05, "loss": 0.0358, "step": 55550 }, { "epoch": 1.1102673754046601, "grad_norm": 0.13730835914611816, "learning_rate": 7.772779489458909e-05, "loss": 0.0373, "step": 55560 }, { "epoch": 1.1104672075456616, "grad_norm": 0.12070710211992264, "learning_rate": 7.771919474565659e-05, "loss": 0.0324, "step": 55570 }, { "epoch": 1.1106670396866631, "grad_norm": 0.1114104762673378, "learning_rate": 7.771059341261103e-05, "loss": 0.0383, "step": 55580 }, { "epoch": 1.1108668718276649, "grad_norm": 0.18264390528202057, "learning_rate": 7.770199089581982e-05, "loss": 0.0335, "step": 55590 }, { "epoch": 1.1110667039686664, "grad_norm": 0.18262435495853424, "learning_rate": 7.769338719565045e-05, "loss": 0.0283, "step": 55600 }, { "epoch": 1.1112665361096679, "grad_norm": 0.11944810301065445, "learning_rate": 7.768478231247044e-05, "loss": 0.047, "step": 55610 }, { "epoch": 1.1114663682506694, "grad_norm": 0.1295057088136673, "learning_rate": 7.767617624664741e-05, "loss": 0.0328, "step": 55620 }, { "epoch": 1.111666200391671, "grad_norm": 0.1128159835934639, "learning_rate": 7.766756899854896e-05, "loss": 0.0358, "step": 55630 }, { "epoch": 1.1118660325326726, "grad_norm": 0.1316627860069275, "learning_rate": 7.76589605685428e-05, "loss": 0.0389, "step": 55640 }, { "epoch": 1.112065864673674, "grad_norm": 0.17108356952667236, "learning_rate": 7.765035095699666e-05, "loss": 0.0343, "step": 55650 }, { "epoch": 1.1122656968146756, "grad_norm": 0.13593319058418274, "learning_rate": 7.764174016427832e-05, "loss": 0.0302, "step": 55660 }, { "epoch": 1.1124655289556773, "grad_norm": 0.15635532140731812, "learning_rate": 7.76331281907556e-05, "loss": 0.0327, "step": 55670 }, { "epoch": 1.1126653610966788, "grad_norm": 0.1306382417678833, "learning_rate": 7.762451503679642e-05, "loss": 0.0378, "step": 55680 }, { "epoch": 1.1128651932376803, "grad_norm": 0.16118377447128296, "learning_rate": 7.761590070276871e-05, "loss": 0.0364, "step": 55690 }, { "epoch": 1.113065025378682, "grad_norm": 0.12263961881399155, "learning_rate": 7.760728518904045e-05, "loss": 0.0295, "step": 55700 }, { "epoch": 1.1132648575196835, "grad_norm": 0.13551048934459686, "learning_rate": 7.759866849597968e-05, "loss": 0.0298, "step": 55710 }, { "epoch": 1.113464689660685, "grad_norm": 0.11022721976041794, "learning_rate": 7.75900506239545e-05, "loss": 0.0264, "step": 55720 }, { "epoch": 1.1136645218016865, "grad_norm": 0.1288880556821823, "learning_rate": 7.758143157333303e-05, "loss": 0.0368, "step": 55730 }, { "epoch": 1.1138643539426882, "grad_norm": 0.10232947021722794, "learning_rate": 7.757281134448348e-05, "loss": 0.0321, "step": 55740 }, { "epoch": 1.1140641860836897, "grad_norm": 0.1407296359539032, "learning_rate": 7.756418993777408e-05, "loss": 0.0262, "step": 55750 }, { "epoch": 1.1142640182246912, "grad_norm": 0.1435806006193161, "learning_rate": 7.755556735357312e-05, "loss": 0.0361, "step": 55760 }, { "epoch": 1.1144638503656927, "grad_norm": 0.10610882192850113, "learning_rate": 7.754694359224895e-05, "loss": 0.0409, "step": 55770 }, { "epoch": 1.1146636825066945, "grad_norm": 0.1179296225309372, "learning_rate": 7.753831865416995e-05, "loss": 0.032, "step": 55780 }, { "epoch": 1.114863514647696, "grad_norm": 0.14878347516059875, "learning_rate": 7.752969253970457e-05, "loss": 0.0405, "step": 55790 }, { "epoch": 1.1150633467886975, "grad_norm": 0.11903699487447739, "learning_rate": 7.752106524922129e-05, "loss": 0.0284, "step": 55800 }, { "epoch": 1.115263178929699, "grad_norm": 0.18350788950920105, "learning_rate": 7.751243678308869e-05, "loss": 0.038, "step": 55810 }, { "epoch": 1.1154630110707007, "grad_norm": 0.13673724234104156, "learning_rate": 7.75038071416753e-05, "loss": 0.0383, "step": 55820 }, { "epoch": 1.1156628432117022, "grad_norm": 0.12228383868932724, "learning_rate": 7.74951763253498e-05, "loss": 0.0352, "step": 55830 }, { "epoch": 1.1158626753527037, "grad_norm": 0.12314848601818085, "learning_rate": 7.748654433448089e-05, "loss": 0.0353, "step": 55840 }, { "epoch": 1.1160625074937054, "grad_norm": 0.13134662806987762, "learning_rate": 7.74779111694373e-05, "loss": 0.0397, "step": 55850 }, { "epoch": 1.116262339634707, "grad_norm": 0.14739012718200684, "learning_rate": 7.746927683058782e-05, "loss": 0.0394, "step": 55860 }, { "epoch": 1.1164621717757084, "grad_norm": 0.15083815157413483, "learning_rate": 7.746064131830129e-05, "loss": 0.0317, "step": 55870 }, { "epoch": 1.1166620039167099, "grad_norm": 0.13654330372810364, "learning_rate": 7.74520046329466e-05, "loss": 0.0419, "step": 55880 }, { "epoch": 1.1168618360577116, "grad_norm": 0.1534428894519806, "learning_rate": 7.744336677489272e-05, "loss": 0.0376, "step": 55890 }, { "epoch": 1.117061668198713, "grad_norm": 0.14589445292949677, "learning_rate": 7.743472774450863e-05, "loss": 0.0311, "step": 55900 }, { "epoch": 1.1172615003397146, "grad_norm": 0.1421247273683548, "learning_rate": 7.742608754216335e-05, "loss": 0.0303, "step": 55910 }, { "epoch": 1.117461332480716, "grad_norm": 0.1364774852991104, "learning_rate": 7.741744616822602e-05, "loss": 0.0307, "step": 55920 }, { "epoch": 1.1176611646217178, "grad_norm": 0.11479643732309341, "learning_rate": 7.740880362306573e-05, "loss": 0.0394, "step": 55930 }, { "epoch": 1.1178609967627193, "grad_norm": 0.10297852009534836, "learning_rate": 7.740015990705172e-05, "loss": 0.0326, "step": 55940 }, { "epoch": 1.1180608289037208, "grad_norm": 0.13382326066493988, "learning_rate": 7.739151502055321e-05, "loss": 0.0283, "step": 55950 }, { "epoch": 1.1182606610447223, "grad_norm": 0.11712388694286346, "learning_rate": 7.738286896393949e-05, "loss": 0.0401, "step": 55960 }, { "epoch": 1.118460493185724, "grad_norm": 0.16353361308574677, "learning_rate": 7.737422173757994e-05, "loss": 0.0338, "step": 55970 }, { "epoch": 1.1186603253267255, "grad_norm": 0.12925036251544952, "learning_rate": 7.736557334184389e-05, "loss": 0.026, "step": 55980 }, { "epoch": 1.118860157467727, "grad_norm": 0.1469002068042755, "learning_rate": 7.735692377710082e-05, "loss": 0.0322, "step": 55990 }, { "epoch": 1.1190599896087288, "grad_norm": 0.15079209208488464, "learning_rate": 7.734827304372025e-05, "loss": 0.0317, "step": 56000 }, { "epoch": 1.1192598217497303, "grad_norm": 0.18669059872627258, "learning_rate": 7.733962114207167e-05, "loss": 0.0318, "step": 56010 }, { "epoch": 1.1194596538907318, "grad_norm": 0.12267027050256729, "learning_rate": 7.733096807252471e-05, "loss": 0.0282, "step": 56020 }, { "epoch": 1.1196594860317333, "grad_norm": 0.18644478917121887, "learning_rate": 7.732231383544902e-05, "loss": 0.0435, "step": 56030 }, { "epoch": 1.119859318172735, "grad_norm": 0.14775234460830688, "learning_rate": 7.731365843121425e-05, "loss": 0.0356, "step": 56040 }, { "epoch": 1.1200591503137365, "grad_norm": 0.14384615421295166, "learning_rate": 7.730500186019018e-05, "loss": 0.0355, "step": 56050 }, { "epoch": 1.120258982454738, "grad_norm": 0.1019844114780426, "learning_rate": 7.729634412274659e-05, "loss": 0.0307, "step": 56060 }, { "epoch": 1.1204588145957395, "grad_norm": 0.1138603538274765, "learning_rate": 7.728768521925332e-05, "loss": 0.0354, "step": 56070 }, { "epoch": 1.1206586467367412, "grad_norm": 0.14866863191127777, "learning_rate": 7.727902515008028e-05, "loss": 0.0404, "step": 56080 }, { "epoch": 1.1208584788777427, "grad_norm": 0.14407142996788025, "learning_rate": 7.727036391559737e-05, "loss": 0.0309, "step": 56090 }, { "epoch": 1.1210583110187442, "grad_norm": 0.11948899179697037, "learning_rate": 7.726170151617464e-05, "loss": 0.0331, "step": 56100 }, { "epoch": 1.121258143159746, "grad_norm": 0.14515896141529083, "learning_rate": 7.725303795218209e-05, "loss": 0.0344, "step": 56110 }, { "epoch": 1.1214579753007474, "grad_norm": 0.14581048488616943, "learning_rate": 7.724437322398982e-05, "loss": 0.0367, "step": 56120 }, { "epoch": 1.121657807441749, "grad_norm": 0.10999849438667297, "learning_rate": 7.723570733196798e-05, "loss": 0.0264, "step": 56130 }, { "epoch": 1.1218576395827504, "grad_norm": 0.13658323884010315, "learning_rate": 7.722704027648676e-05, "loss": 0.0284, "step": 56140 }, { "epoch": 1.1220574717237521, "grad_norm": 0.13606242835521698, "learning_rate": 7.72183720579164e-05, "loss": 0.038, "step": 56150 }, { "epoch": 1.1222573038647536, "grad_norm": 0.1439540982246399, "learning_rate": 7.720970267662717e-05, "loss": 0.0288, "step": 56160 }, { "epoch": 1.1224571360057551, "grad_norm": 0.1270395964384079, "learning_rate": 7.720103213298945e-05, "loss": 0.0336, "step": 56170 }, { "epoch": 1.1226569681467566, "grad_norm": 0.10710359364748001, "learning_rate": 7.719236042737359e-05, "loss": 0.0373, "step": 56180 }, { "epoch": 1.1228568002877584, "grad_norm": 0.166272833943367, "learning_rate": 7.718368756015006e-05, "loss": 0.0394, "step": 56190 }, { "epoch": 1.1230566324287599, "grad_norm": 0.16930241882801056, "learning_rate": 7.717501353168932e-05, "loss": 0.0417, "step": 56200 }, { "epoch": 1.1232564645697614, "grad_norm": 0.1426674872636795, "learning_rate": 7.716633834236193e-05, "loss": 0.0321, "step": 56210 }, { "epoch": 1.1234562967107629, "grad_norm": 0.18395720422267914, "learning_rate": 7.715766199253848e-05, "loss": 0.0319, "step": 56220 }, { "epoch": 1.1236561288517646, "grad_norm": 0.15842987596988678, "learning_rate": 7.71489844825896e-05, "loss": 0.0384, "step": 56230 }, { "epoch": 1.123855960992766, "grad_norm": 0.13384941220283508, "learning_rate": 7.714030581288596e-05, "loss": 0.0327, "step": 56240 }, { "epoch": 1.1240557931337676, "grad_norm": 0.13312748074531555, "learning_rate": 7.713162598379833e-05, "loss": 0.0288, "step": 56250 }, { "epoch": 1.124255625274769, "grad_norm": 0.10040907561779022, "learning_rate": 7.712294499569748e-05, "loss": 0.0301, "step": 56260 }, { "epoch": 1.1244554574157708, "grad_norm": 0.11914680153131485, "learning_rate": 7.711426284895424e-05, "loss": 0.0264, "step": 56270 }, { "epoch": 1.1246552895567723, "grad_norm": 0.18596866726875305, "learning_rate": 7.71055795439395e-05, "loss": 0.0273, "step": 56280 }, { "epoch": 1.1248551216977738, "grad_norm": 0.1264795958995819, "learning_rate": 7.70968950810242e-05, "loss": 0.0318, "step": 56290 }, { "epoch": 1.1250549538387755, "grad_norm": 0.1416316032409668, "learning_rate": 7.708820946057932e-05, "loss": 0.0286, "step": 56300 }, { "epoch": 1.125254785979777, "grad_norm": 0.13297387957572937, "learning_rate": 7.707952268297589e-05, "loss": 0.0274, "step": 56310 }, { "epoch": 1.1254546181207785, "grad_norm": 0.15359826385974884, "learning_rate": 7.707083474858501e-05, "loss": 0.0372, "step": 56320 }, { "epoch": 1.12565445026178, "grad_norm": 0.16178907454013824, "learning_rate": 7.70621456577778e-05, "loss": 0.0351, "step": 56330 }, { "epoch": 1.1258542824027817, "grad_norm": 0.11512299627065659, "learning_rate": 7.705345541092544e-05, "loss": 0.0315, "step": 56340 }, { "epoch": 1.1260541145437832, "grad_norm": 0.11726215481758118, "learning_rate": 7.704476400839915e-05, "loss": 0.0256, "step": 56350 }, { "epoch": 1.1262539466847847, "grad_norm": 0.10807417333126068, "learning_rate": 7.703607145057023e-05, "loss": 0.0376, "step": 56360 }, { "epoch": 1.1264537788257862, "grad_norm": 0.13967935740947723, "learning_rate": 7.702737773781002e-05, "loss": 0.0288, "step": 56370 }, { "epoch": 1.126653610966788, "grad_norm": 0.1548096388578415, "learning_rate": 7.701868287048985e-05, "loss": 0.027, "step": 56380 }, { "epoch": 1.1268534431077895, "grad_norm": 0.14483991265296936, "learning_rate": 7.700998684898122e-05, "loss": 0.0351, "step": 56390 }, { "epoch": 1.127053275248791, "grad_norm": 0.13319073617458344, "learning_rate": 7.700128967365556e-05, "loss": 0.0375, "step": 56400 }, { "epoch": 1.1272531073897927, "grad_norm": 0.1333056390285492, "learning_rate": 7.69925913448844e-05, "loss": 0.0366, "step": 56410 }, { "epoch": 1.1274529395307942, "grad_norm": 0.1290370672941208, "learning_rate": 7.698389186303933e-05, "loss": 0.0326, "step": 56420 }, { "epoch": 1.1276527716717957, "grad_norm": 0.11836012452840805, "learning_rate": 7.697519122849197e-05, "loss": 0.0282, "step": 56430 }, { "epoch": 1.1278526038127972, "grad_norm": 0.14630639553070068, "learning_rate": 7.6966489441614e-05, "loss": 0.0439, "step": 56440 }, { "epoch": 1.128052435953799, "grad_norm": 0.1214488223195076, "learning_rate": 7.695778650277713e-05, "loss": 0.0318, "step": 56450 }, { "epoch": 1.1282522680948004, "grad_norm": 0.2002963274717331, "learning_rate": 7.694908241235314e-05, "loss": 0.0401, "step": 56460 }, { "epoch": 1.128452100235802, "grad_norm": 0.13035982847213745, "learning_rate": 7.694037717071387e-05, "loss": 0.0334, "step": 56470 }, { "epoch": 1.1286519323768034, "grad_norm": 0.11393220722675323, "learning_rate": 7.693167077823118e-05, "loss": 0.032, "step": 56480 }, { "epoch": 1.1288517645178051, "grad_norm": 0.1547337770462036, "learning_rate": 7.692296323527697e-05, "loss": 0.0394, "step": 56490 }, { "epoch": 1.1290515966588066, "grad_norm": 0.13978831470012665, "learning_rate": 7.691425454222325e-05, "loss": 0.033, "step": 56500 }, { "epoch": 1.129251428799808, "grad_norm": 0.15118683874607086, "learning_rate": 7.690554469944201e-05, "loss": 0.0409, "step": 56510 }, { "epoch": 1.1294512609408098, "grad_norm": 0.1191638708114624, "learning_rate": 7.689683370730533e-05, "loss": 0.0332, "step": 56520 }, { "epoch": 1.1296510930818113, "grad_norm": 0.12779614329338074, "learning_rate": 7.688812156618533e-05, "loss": 0.034, "step": 56530 }, { "epoch": 1.1298509252228128, "grad_norm": 0.09626220166683197, "learning_rate": 7.687940827645416e-05, "loss": 0.0352, "step": 56540 }, { "epoch": 1.1300507573638143, "grad_norm": 0.16521196067333221, "learning_rate": 7.687069383848406e-05, "loss": 0.0299, "step": 56550 }, { "epoch": 1.130250589504816, "grad_norm": 0.12657122313976288, "learning_rate": 7.686197825264727e-05, "loss": 0.0314, "step": 56560 }, { "epoch": 1.1304504216458175, "grad_norm": 0.15894271433353424, "learning_rate": 7.685326151931612e-05, "loss": 0.0282, "step": 56570 }, { "epoch": 1.130650253786819, "grad_norm": 0.13067306578159332, "learning_rate": 7.684454363886297e-05, "loss": 0.0315, "step": 56580 }, { "epoch": 1.1308500859278205, "grad_norm": 0.1099950522184372, "learning_rate": 7.683582461166024e-05, "loss": 0.0284, "step": 56590 }, { "epoch": 1.1310499180688223, "grad_norm": 0.13464166224002838, "learning_rate": 7.682710443808036e-05, "loss": 0.0343, "step": 56600 }, { "epoch": 1.1312497502098238, "grad_norm": 0.12051894515752792, "learning_rate": 7.681838311849587e-05, "loss": 0.0294, "step": 56610 }, { "epoch": 1.1314495823508253, "grad_norm": 0.16018785536289215, "learning_rate": 7.680966065327933e-05, "loss": 0.031, "step": 56620 }, { "epoch": 1.131649414491827, "grad_norm": 0.15333114564418793, "learning_rate": 7.680093704280333e-05, "loss": 0.0364, "step": 56630 }, { "epoch": 1.1318492466328285, "grad_norm": 0.11735818535089493, "learning_rate": 7.679221228744054e-05, "loss": 0.0313, "step": 56640 }, { "epoch": 1.13204907877383, "grad_norm": 0.2022658884525299, "learning_rate": 7.678348638756365e-05, "loss": 0.0466, "step": 56650 }, { "epoch": 1.1322489109148315, "grad_norm": 0.13587529957294464, "learning_rate": 7.677475934354543e-05, "loss": 0.0276, "step": 56660 }, { "epoch": 1.132448743055833, "grad_norm": 0.12007590383291245, "learning_rate": 7.67660311557587e-05, "loss": 0.0329, "step": 56670 }, { "epoch": 1.1326485751968347, "grad_norm": 0.1373937875032425, "learning_rate": 7.675730182457626e-05, "loss": 0.0294, "step": 56680 }, { "epoch": 1.1328484073378362, "grad_norm": 0.13779741525650024, "learning_rate": 7.674857135037104e-05, "loss": 0.0326, "step": 56690 }, { "epoch": 1.1330482394788377, "grad_norm": 0.14557033777236938, "learning_rate": 7.673983973351601e-05, "loss": 0.0379, "step": 56700 }, { "epoch": 1.1332480716198394, "grad_norm": 0.1326465904712677, "learning_rate": 7.673110697438412e-05, "loss": 0.0323, "step": 56710 }, { "epoch": 1.133447903760841, "grad_norm": 0.13646261394023895, "learning_rate": 7.672237307334847e-05, "loss": 0.0329, "step": 56720 }, { "epoch": 1.1336477359018424, "grad_norm": 0.11574268341064453, "learning_rate": 7.671363803078215e-05, "loss": 0.0387, "step": 56730 }, { "epoch": 1.133847568042844, "grad_norm": 0.10514404624700546, "learning_rate": 7.670490184705826e-05, "loss": 0.0289, "step": 56740 }, { "epoch": 1.1340474001838456, "grad_norm": 0.14641040563583374, "learning_rate": 7.669616452255001e-05, "loss": 0.0313, "step": 56750 }, { "epoch": 1.1342472323248471, "grad_norm": 0.1397874802350998, "learning_rate": 7.668742605763067e-05, "loss": 0.0362, "step": 56760 }, { "epoch": 1.1344470644658486, "grad_norm": 0.11920182406902313, "learning_rate": 7.667868645267353e-05, "loss": 0.031, "step": 56770 }, { "epoch": 1.1346468966068501, "grad_norm": 0.1429857611656189, "learning_rate": 7.66699457080519e-05, "loss": 0.0356, "step": 56780 }, { "epoch": 1.1348467287478519, "grad_norm": 0.12546341121196747, "learning_rate": 7.666120382413918e-05, "loss": 0.041, "step": 56790 }, { "epoch": 1.1350465608888534, "grad_norm": 0.13728220760822296, "learning_rate": 7.665246080130884e-05, "loss": 0.0259, "step": 56800 }, { "epoch": 1.1352463930298549, "grad_norm": 0.1620642989873886, "learning_rate": 7.664371663993428e-05, "loss": 0.037, "step": 56810 }, { "epoch": 1.1354462251708566, "grad_norm": 0.11880146712064743, "learning_rate": 7.663497134038916e-05, "loss": 0.0362, "step": 56820 }, { "epoch": 1.135646057311858, "grad_norm": 0.15947571396827698, "learning_rate": 7.662622490304695e-05, "loss": 0.039, "step": 56830 }, { "epoch": 1.1358458894528596, "grad_norm": 0.15923677384853363, "learning_rate": 7.661747732828133e-05, "loss": 0.0321, "step": 56840 }, { "epoch": 1.136045721593861, "grad_norm": 0.16243477165699005, "learning_rate": 7.660872861646597e-05, "loss": 0.0345, "step": 56850 }, { "epoch": 1.1362455537348628, "grad_norm": 0.12041797488927841, "learning_rate": 7.659997876797462e-05, "loss": 0.0301, "step": 56860 }, { "epoch": 1.1364453858758643, "grad_norm": 0.15739507973194122, "learning_rate": 7.659122778318104e-05, "loss": 0.0325, "step": 56870 }, { "epoch": 1.1366452180168658, "grad_norm": 0.11641785502433777, "learning_rate": 7.658247566245907e-05, "loss": 0.0418, "step": 56880 }, { "epoch": 1.1368450501578673, "grad_norm": 0.1553041636943817, "learning_rate": 7.657372240618255e-05, "loss": 0.0298, "step": 56890 }, { "epoch": 1.137044882298869, "grad_norm": 0.10717964917421341, "learning_rate": 7.656496801472543e-05, "loss": 0.0301, "step": 56900 }, { "epoch": 1.1372447144398705, "grad_norm": 0.19292834401130676, "learning_rate": 7.65562124884617e-05, "loss": 0.04, "step": 56910 }, { "epoch": 1.137444546580872, "grad_norm": 0.16040199995040894, "learning_rate": 7.654745582776533e-05, "loss": 0.0343, "step": 56920 }, { "epoch": 1.1376443787218737, "grad_norm": 0.13158738613128662, "learning_rate": 7.653869803301044e-05, "loss": 0.0337, "step": 56930 }, { "epoch": 1.1378442108628752, "grad_norm": 0.1610623002052307, "learning_rate": 7.65299391045711e-05, "loss": 0.0322, "step": 56940 }, { "epoch": 1.1380440430038767, "grad_norm": 0.15476855635643005, "learning_rate": 7.652117904282152e-05, "loss": 0.0417, "step": 56950 }, { "epoch": 1.1382438751448782, "grad_norm": 0.12547069787979126, "learning_rate": 7.651241784813588e-05, "loss": 0.0334, "step": 56960 }, { "epoch": 1.13844370728588, "grad_norm": 0.13201674818992615, "learning_rate": 7.650365552088845e-05, "loss": 0.0357, "step": 56970 }, { "epoch": 1.1386435394268815, "grad_norm": 0.17892397940158844, "learning_rate": 7.649489206145356e-05, "loss": 0.0434, "step": 56980 }, { "epoch": 1.138843371567883, "grad_norm": 0.13669615983963013, "learning_rate": 7.648612747020553e-05, "loss": 0.0352, "step": 56990 }, { "epoch": 1.1390432037088845, "grad_norm": 0.124549500644207, "learning_rate": 7.647736174751883e-05, "loss": 0.0385, "step": 57000 }, { "epoch": 1.1392430358498862, "grad_norm": 0.1243499293923378, "learning_rate": 7.646859489376785e-05, "loss": 0.0303, "step": 57010 }, { "epoch": 1.1394428679908877, "grad_norm": 0.14600643515586853, "learning_rate": 7.645982690932713e-05, "loss": 0.034, "step": 57020 }, { "epoch": 1.1396427001318892, "grad_norm": 0.19189158082008362, "learning_rate": 7.64510577945712e-05, "loss": 0.0345, "step": 57030 }, { "epoch": 1.139842532272891, "grad_norm": 0.13680441677570343, "learning_rate": 7.644228754987468e-05, "loss": 0.037, "step": 57040 }, { "epoch": 1.1400423644138924, "grad_norm": 0.13258811831474304, "learning_rate": 7.643351617561221e-05, "loss": 0.0357, "step": 57050 }, { "epoch": 1.140242196554894, "grad_norm": 0.14713259041309357, "learning_rate": 7.64247436721585e-05, "loss": 0.0288, "step": 57060 }, { "epoch": 1.1404420286958954, "grad_norm": 0.13227452337741852, "learning_rate": 7.641597003988828e-05, "loss": 0.0372, "step": 57070 }, { "epoch": 1.140641860836897, "grad_norm": 0.12818780541419983, "learning_rate": 7.640719527917634e-05, "loss": 0.0389, "step": 57080 }, { "epoch": 1.1408416929778986, "grad_norm": 0.13711017370224, "learning_rate": 7.639841939039754e-05, "loss": 0.0346, "step": 57090 }, { "epoch": 1.1410415251189001, "grad_norm": 0.12538008391857147, "learning_rate": 7.638964237392675e-05, "loss": 0.0345, "step": 57100 }, { "epoch": 1.1412413572599016, "grad_norm": 0.1153169497847557, "learning_rate": 7.63808642301389e-05, "loss": 0.0275, "step": 57110 }, { "epoch": 1.1414411894009033, "grad_norm": 0.1455181986093521, "learning_rate": 7.637208495940902e-05, "loss": 0.0272, "step": 57120 }, { "epoch": 1.1416410215419048, "grad_norm": 0.1476776897907257, "learning_rate": 7.636330456211212e-05, "loss": 0.0336, "step": 57130 }, { "epoch": 1.1418408536829063, "grad_norm": 0.13660912215709686, "learning_rate": 7.635452303862324e-05, "loss": 0.0242, "step": 57140 }, { "epoch": 1.1420406858239078, "grad_norm": 0.09673433005809784, "learning_rate": 7.63457403893176e-05, "loss": 0.0394, "step": 57150 }, { "epoch": 1.1422405179649096, "grad_norm": 0.12296269088983536, "learning_rate": 7.633695661457031e-05, "loss": 0.0312, "step": 57160 }, { "epoch": 1.142440350105911, "grad_norm": 0.12929300963878632, "learning_rate": 7.632817171475661e-05, "loss": 0.0354, "step": 57170 }, { "epoch": 1.1426401822469126, "grad_norm": 0.10366176068782806, "learning_rate": 7.631938569025179e-05, "loss": 0.0291, "step": 57180 }, { "epoch": 1.142840014387914, "grad_norm": 0.12801504135131836, "learning_rate": 7.631059854143116e-05, "loss": 0.0383, "step": 57190 }, { "epoch": 1.1430398465289158, "grad_norm": 0.12911760807037354, "learning_rate": 7.63018102686701e-05, "loss": 0.0344, "step": 57200 }, { "epoch": 1.1432396786699173, "grad_norm": 0.12125732004642487, "learning_rate": 7.6293020872344e-05, "loss": 0.0346, "step": 57210 }, { "epoch": 1.1434395108109188, "grad_norm": 0.1231606975197792, "learning_rate": 7.628423035282838e-05, "loss": 0.0305, "step": 57220 }, { "epoch": 1.1436393429519205, "grad_norm": 0.16040632128715515, "learning_rate": 7.627543871049871e-05, "loss": 0.0446, "step": 57230 }, { "epoch": 1.143839175092922, "grad_norm": 0.12438897788524628, "learning_rate": 7.626664594573057e-05, "loss": 0.0373, "step": 57240 }, { "epoch": 1.1440390072339235, "grad_norm": 0.09412486106157303, "learning_rate": 7.625785205889958e-05, "loss": 0.035, "step": 57250 }, { "epoch": 1.144238839374925, "grad_norm": 0.14710983633995056, "learning_rate": 7.624905705038137e-05, "loss": 0.0345, "step": 57260 }, { "epoch": 1.1444386715159267, "grad_norm": 0.158950537443161, "learning_rate": 7.624026092055166e-05, "loss": 0.0307, "step": 57270 }, { "epoch": 1.1446385036569282, "grad_norm": 0.11690382659435272, "learning_rate": 7.623146366978622e-05, "loss": 0.0385, "step": 57280 }, { "epoch": 1.1448383357979297, "grad_norm": 0.1232309639453888, "learning_rate": 7.622266529846084e-05, "loss": 0.0316, "step": 57290 }, { "epoch": 1.1450381679389312, "grad_norm": 0.1810358762741089, "learning_rate": 7.621386580695137e-05, "loss": 0.0305, "step": 57300 }, { "epoch": 1.145238000079933, "grad_norm": 0.14700745046138763, "learning_rate": 7.62050651956337e-05, "loss": 0.027, "step": 57310 }, { "epoch": 1.1454378322209344, "grad_norm": 0.11671201139688492, "learning_rate": 7.61962634648838e-05, "loss": 0.0333, "step": 57320 }, { "epoch": 1.145637664361936, "grad_norm": 0.13327324390411377, "learning_rate": 7.618746061507765e-05, "loss": 0.0406, "step": 57330 }, { "epoch": 1.1458374965029376, "grad_norm": 0.1323855072259903, "learning_rate": 7.617865664659128e-05, "loss": 0.0364, "step": 57340 }, { "epoch": 1.1460373286439391, "grad_norm": 0.1011352688074112, "learning_rate": 7.616985155980079e-05, "loss": 0.0381, "step": 57350 }, { "epoch": 1.1462371607849406, "grad_norm": 0.10385138541460037, "learning_rate": 7.616104535508232e-05, "loss": 0.033, "step": 57360 }, { "epoch": 1.1464369929259421, "grad_norm": 0.12986475229263306, "learning_rate": 7.615223803281206e-05, "loss": 0.035, "step": 57370 }, { "epoch": 1.1466368250669436, "grad_norm": 0.106245256960392, "learning_rate": 7.614342959336622e-05, "loss": 0.0308, "step": 57380 }, { "epoch": 1.1468366572079454, "grad_norm": 0.1291751116514206, "learning_rate": 7.61346200371211e-05, "loss": 0.0329, "step": 57390 }, { "epoch": 1.1470364893489469, "grad_norm": 0.1389077752828598, "learning_rate": 7.612580936445304e-05, "loss": 0.0316, "step": 57400 }, { "epoch": 1.1472363214899484, "grad_norm": 0.18829353153705597, "learning_rate": 7.611699757573835e-05, "loss": 0.0314, "step": 57410 }, { "epoch": 1.14743615363095, "grad_norm": 0.1550585925579071, "learning_rate": 7.610818467135355e-05, "loss": 0.0323, "step": 57420 }, { "epoch": 1.1476359857719516, "grad_norm": 0.1358419507741928, "learning_rate": 7.609937065167505e-05, "loss": 0.0328, "step": 57430 }, { "epoch": 1.147835817912953, "grad_norm": 0.13547006249427795, "learning_rate": 7.609055551707937e-05, "loss": 0.0357, "step": 57440 }, { "epoch": 1.1480356500539546, "grad_norm": 0.10745007544755936, "learning_rate": 7.608173926794312e-05, "loss": 0.035, "step": 57450 }, { "epoch": 1.1482354821949563, "grad_norm": 0.15527474880218506, "learning_rate": 7.607292190464285e-05, "loss": 0.0306, "step": 57460 }, { "epoch": 1.1484353143359578, "grad_norm": 0.12242557853460312, "learning_rate": 7.606410342755527e-05, "loss": 0.0298, "step": 57470 }, { "epoch": 1.1486351464769593, "grad_norm": 0.184173122048378, "learning_rate": 7.605528383705706e-05, "loss": 0.0297, "step": 57480 }, { "epoch": 1.1488349786179608, "grad_norm": 0.13838841021060944, "learning_rate": 7.604646313352499e-05, "loss": 0.0286, "step": 57490 }, { "epoch": 1.1490348107589625, "grad_norm": 0.1621946096420288, "learning_rate": 7.603764131733588e-05, "loss": 0.0321, "step": 57500 }, { "epoch": 1.149234642899964, "grad_norm": 0.1428556889295578, "learning_rate": 7.602881838886655e-05, "loss": 0.0372, "step": 57510 }, { "epoch": 1.1494344750409655, "grad_norm": 0.17786549031734467, "learning_rate": 7.601999434849392e-05, "loss": 0.042, "step": 57520 }, { "epoch": 1.1496343071819672, "grad_norm": 0.12380992621183395, "learning_rate": 7.601116919659494e-05, "loss": 0.0355, "step": 57530 }, { "epoch": 1.1498341393229687, "grad_norm": 0.14350199699401855, "learning_rate": 7.600234293354659e-05, "loss": 0.0353, "step": 57540 }, { "epoch": 1.1500339714639702, "grad_norm": 0.1396697610616684, "learning_rate": 7.599351555972591e-05, "loss": 0.0394, "step": 57550 }, { "epoch": 1.1502338036049717, "grad_norm": 0.14393170177936554, "learning_rate": 7.598468707551001e-05, "loss": 0.0398, "step": 57560 }, { "epoch": 1.1504336357459735, "grad_norm": 0.1008683368563652, "learning_rate": 7.5975857481276e-05, "loss": 0.029, "step": 57570 }, { "epoch": 1.150633467886975, "grad_norm": 0.13640613853931427, "learning_rate": 7.596702677740108e-05, "loss": 0.0375, "step": 57580 }, { "epoch": 1.1508333000279765, "grad_norm": 0.12679824233055115, "learning_rate": 7.595819496426249e-05, "loss": 0.0322, "step": 57590 }, { "epoch": 1.151033132168978, "grad_norm": 0.17222660779953003, "learning_rate": 7.59493620422375e-05, "loss": 0.0343, "step": 57600 }, { "epoch": 1.1512329643099797, "grad_norm": 0.1185910701751709, "learning_rate": 7.59405280117034e-05, "loss": 0.0319, "step": 57610 }, { "epoch": 1.1514327964509812, "grad_norm": 0.15287557244300842, "learning_rate": 7.593169287303763e-05, "loss": 0.0313, "step": 57620 }, { "epoch": 1.1516326285919827, "grad_norm": 0.14654450118541718, "learning_rate": 7.592285662661758e-05, "loss": 0.0292, "step": 57630 }, { "epoch": 1.1518324607329844, "grad_norm": 0.13293328881263733, "learning_rate": 7.59140192728207e-05, "loss": 0.0335, "step": 57640 }, { "epoch": 1.152032292873986, "grad_norm": 0.12095653265714645, "learning_rate": 7.590518081202453e-05, "loss": 0.0238, "step": 57650 }, { "epoch": 1.1522321250149874, "grad_norm": 0.131158247590065, "learning_rate": 7.589634124460663e-05, "loss": 0.0291, "step": 57660 }, { "epoch": 1.152431957155989, "grad_norm": 0.11864019185304642, "learning_rate": 7.58875005709446e-05, "loss": 0.034, "step": 57670 }, { "epoch": 1.1526317892969906, "grad_norm": 0.1369498074054718, "learning_rate": 7.587865879141611e-05, "loss": 0.0275, "step": 57680 }, { "epoch": 1.1528316214379921, "grad_norm": 0.12331096082925797, "learning_rate": 7.586981590639883e-05, "loss": 0.0382, "step": 57690 }, { "epoch": 1.1530314535789936, "grad_norm": 0.1472320854663849, "learning_rate": 7.586097191627058e-05, "loss": 0.0352, "step": 57700 }, { "epoch": 1.1532312857199951, "grad_norm": 0.14821690320968628, "learning_rate": 7.585212682140908e-05, "loss": 0.0366, "step": 57710 }, { "epoch": 1.1534311178609968, "grad_norm": 0.15315300226211548, "learning_rate": 7.584328062219223e-05, "loss": 0.0313, "step": 57720 }, { "epoch": 1.1536309500019983, "grad_norm": 0.14542458951473236, "learning_rate": 7.58344333189979e-05, "loss": 0.0366, "step": 57730 }, { "epoch": 1.1538307821429998, "grad_norm": 0.17033648490905762, "learning_rate": 7.582558491220407e-05, "loss": 0.046, "step": 57740 }, { "epoch": 1.1540306142840016, "grad_norm": 0.14173895120620728, "learning_rate": 7.581673540218866e-05, "loss": 0.0359, "step": 57750 }, { "epoch": 1.154230446425003, "grad_norm": 0.12498294562101364, "learning_rate": 7.580788478932975e-05, "loss": 0.0276, "step": 57760 }, { "epoch": 1.1544302785660046, "grad_norm": 0.1321049928665161, "learning_rate": 7.579903307400542e-05, "loss": 0.0286, "step": 57770 }, { "epoch": 1.154630110707006, "grad_norm": 0.12735196948051453, "learning_rate": 7.579018025659378e-05, "loss": 0.0349, "step": 57780 }, { "epoch": 1.1548299428480076, "grad_norm": 0.1693991720676422, "learning_rate": 7.578132633747301e-05, "loss": 0.0398, "step": 57790 }, { "epoch": 1.1550297749890093, "grad_norm": 0.1254742294549942, "learning_rate": 7.577247131702136e-05, "loss": 0.0332, "step": 57800 }, { "epoch": 1.1552296071300108, "grad_norm": 0.14543774724006653, "learning_rate": 7.576361519561709e-05, "loss": 0.0297, "step": 57810 }, { "epoch": 1.1554294392710123, "grad_norm": 0.15218719840049744, "learning_rate": 7.575475797363849e-05, "loss": 0.035, "step": 57820 }, { "epoch": 1.155629271412014, "grad_norm": 0.1769212931394577, "learning_rate": 7.574589965146395e-05, "loss": 0.0315, "step": 57830 }, { "epoch": 1.1558291035530155, "grad_norm": 0.15943770110607147, "learning_rate": 7.573704022947185e-05, "loss": 0.04, "step": 57840 }, { "epoch": 1.156028935694017, "grad_norm": 0.15248340368270874, "learning_rate": 7.572817970804072e-05, "loss": 0.0426, "step": 57850 }, { "epoch": 1.1562287678350185, "grad_norm": 0.1151675209403038, "learning_rate": 7.571931808754896e-05, "loss": 0.0343, "step": 57860 }, { "epoch": 1.1564285999760202, "grad_norm": 0.18916262686252594, "learning_rate": 7.571045536837523e-05, "loss": 0.0262, "step": 57870 }, { "epoch": 1.1566284321170217, "grad_norm": 0.1261109560728073, "learning_rate": 7.570159155089804e-05, "loss": 0.0268, "step": 57880 }, { "epoch": 1.1568282642580232, "grad_norm": 0.13102291524410248, "learning_rate": 7.56927266354961e-05, "loss": 0.0285, "step": 57890 }, { "epoch": 1.1570280963990247, "grad_norm": 0.12091229110956192, "learning_rate": 7.568386062254807e-05, "loss": 0.0312, "step": 57900 }, { "epoch": 1.1572279285400264, "grad_norm": 0.09561894834041595, "learning_rate": 7.567499351243268e-05, "loss": 0.0225, "step": 57910 }, { "epoch": 1.157427760681028, "grad_norm": 0.14023031294345856, "learning_rate": 7.566612530552876e-05, "loss": 0.0389, "step": 57920 }, { "epoch": 1.1576275928220294, "grad_norm": 0.1561499834060669, "learning_rate": 7.56572560022151e-05, "loss": 0.0312, "step": 57930 }, { "epoch": 1.1578274249630311, "grad_norm": 0.12849803268909454, "learning_rate": 7.56483856028706e-05, "loss": 0.0371, "step": 57940 }, { "epoch": 1.1580272571040326, "grad_norm": 0.14024437963962555, "learning_rate": 7.563951410787418e-05, "loss": 0.0321, "step": 57950 }, { "epoch": 1.1582270892450341, "grad_norm": 0.1221591904759407, "learning_rate": 7.563064151760483e-05, "loss": 0.0334, "step": 57960 }, { "epoch": 1.1584269213860356, "grad_norm": 0.14234209060668945, "learning_rate": 7.562176783244155e-05, "loss": 0.0349, "step": 57970 }, { "epoch": 1.1586267535270374, "grad_norm": 0.15128226578235626, "learning_rate": 7.561289305276342e-05, "loss": 0.0427, "step": 57980 }, { "epoch": 1.1588265856680389, "grad_norm": 0.09286532551050186, "learning_rate": 7.560401717894956e-05, "loss": 0.0298, "step": 57990 }, { "epoch": 1.1590264178090404, "grad_norm": 0.14023208618164062, "learning_rate": 7.559514021137913e-05, "loss": 0.0329, "step": 58000 }, { "epoch": 1.1592262499500419, "grad_norm": 0.13233302533626556, "learning_rate": 7.558626215043129e-05, "loss": 0.0456, "step": 58010 }, { "epoch": 1.1594260820910436, "grad_norm": 0.11785855889320374, "learning_rate": 7.557738299648537e-05, "loss": 0.0281, "step": 58020 }, { "epoch": 1.159625914232045, "grad_norm": 0.13271771371364594, "learning_rate": 7.556850274992063e-05, "loss": 0.0399, "step": 58030 }, { "epoch": 1.1598257463730466, "grad_norm": 0.0946059450507164, "learning_rate": 7.555962141111641e-05, "loss": 0.0279, "step": 58040 }, { "epoch": 1.1600255785140483, "grad_norm": 0.12759774923324585, "learning_rate": 7.555073898045213e-05, "loss": 0.0286, "step": 58050 }, { "epoch": 1.1602254106550498, "grad_norm": 0.12185816466808319, "learning_rate": 7.55418554583072e-05, "loss": 0.0269, "step": 58060 }, { "epoch": 1.1604252427960513, "grad_norm": 0.13506071269512177, "learning_rate": 7.553297084506114e-05, "loss": 0.0375, "step": 58070 }, { "epoch": 1.1606250749370528, "grad_norm": 0.12402044236660004, "learning_rate": 7.552408514109348e-05, "loss": 0.0395, "step": 58080 }, { "epoch": 1.1608249070780545, "grad_norm": 0.16777515411376953, "learning_rate": 7.551519834678379e-05, "loss": 0.0345, "step": 58090 }, { "epoch": 1.161024739219056, "grad_norm": 0.125700905919075, "learning_rate": 7.55063104625117e-05, "loss": 0.0277, "step": 58100 }, { "epoch": 1.1612245713600575, "grad_norm": 0.12181878834962845, "learning_rate": 7.549742148865686e-05, "loss": 0.036, "step": 58110 }, { "epoch": 1.161424403501059, "grad_norm": 0.15520718693733215, "learning_rate": 7.548853142559904e-05, "loss": 0.0339, "step": 58120 }, { "epoch": 1.1616242356420607, "grad_norm": 0.11593442410230637, "learning_rate": 7.547964027371798e-05, "loss": 0.0362, "step": 58130 }, { "epoch": 1.1618240677830622, "grad_norm": 0.1982261836528778, "learning_rate": 7.54707480333935e-05, "loss": 0.0325, "step": 58140 }, { "epoch": 1.1620238999240637, "grad_norm": 0.12851999700069427, "learning_rate": 7.546185470500543e-05, "loss": 0.0267, "step": 58150 }, { "epoch": 1.1622237320650655, "grad_norm": 0.15294799208641052, "learning_rate": 7.545296028893372e-05, "loss": 0.036, "step": 58160 }, { "epoch": 1.162423564206067, "grad_norm": 0.14103686809539795, "learning_rate": 7.54440647855583e-05, "loss": 0.0425, "step": 58170 }, { "epoch": 1.1626233963470685, "grad_norm": 0.11732355505228043, "learning_rate": 7.543516819525916e-05, "loss": 0.0368, "step": 58180 }, { "epoch": 1.16282322848807, "grad_norm": 0.13271541893482208, "learning_rate": 7.54262705184164e-05, "loss": 0.0342, "step": 58190 }, { "epoch": 1.1630230606290715, "grad_norm": 0.14647717773914337, "learning_rate": 7.541737175541004e-05, "loss": 0.0328, "step": 58200 }, { "epoch": 1.1632228927700732, "grad_norm": 0.18492721021175385, "learning_rate": 7.540847190662026e-05, "loss": 0.0476, "step": 58210 }, { "epoch": 1.1634227249110747, "grad_norm": 0.11704647541046143, "learning_rate": 7.539957097242723e-05, "loss": 0.0314, "step": 58220 }, { "epoch": 1.1636225570520762, "grad_norm": 0.14009332656860352, "learning_rate": 7.53906689532112e-05, "loss": 0.03, "step": 58230 }, { "epoch": 1.163822389193078, "grad_norm": 0.11790347844362259, "learning_rate": 7.538176584935244e-05, "loss": 0.0328, "step": 58240 }, { "epoch": 1.1640222213340794, "grad_norm": 0.1259237378835678, "learning_rate": 7.537286166123127e-05, "loss": 0.0319, "step": 58250 }, { "epoch": 1.164222053475081, "grad_norm": 0.14075812697410583, "learning_rate": 7.536395638922804e-05, "loss": 0.0266, "step": 58260 }, { "epoch": 1.1644218856160824, "grad_norm": 0.16327762603759766, "learning_rate": 7.53550500337232e-05, "loss": 0.0388, "step": 58270 }, { "epoch": 1.1646217177570841, "grad_norm": 0.13621322810649872, "learning_rate": 7.534614259509721e-05, "loss": 0.0311, "step": 58280 }, { "epoch": 1.1648215498980856, "grad_norm": 0.12177380919456482, "learning_rate": 7.533723407373057e-05, "loss": 0.0299, "step": 58290 }, { "epoch": 1.1650213820390871, "grad_norm": 0.1531553864479065, "learning_rate": 7.532832447000384e-05, "loss": 0.0336, "step": 58300 }, { "epoch": 1.1652212141800886, "grad_norm": 0.11238230019807816, "learning_rate": 7.531941378429759e-05, "loss": 0.0311, "step": 58310 }, { "epoch": 1.1654210463210903, "grad_norm": 0.10172143578529358, "learning_rate": 7.531050201699252e-05, "loss": 0.0332, "step": 58320 }, { "epoch": 1.1656208784620918, "grad_norm": 0.1106031984090805, "learning_rate": 7.53015891684693e-05, "loss": 0.0281, "step": 58330 }, { "epoch": 1.1658207106030933, "grad_norm": 0.1099640503525734, "learning_rate": 7.529267523910867e-05, "loss": 0.0341, "step": 58340 }, { "epoch": 1.166020542744095, "grad_norm": 0.1307123303413391, "learning_rate": 7.528376022929142e-05, "loss": 0.0372, "step": 58350 }, { "epoch": 1.1662203748850966, "grad_norm": 0.15093515813350677, "learning_rate": 7.527484413939835e-05, "loss": 0.0352, "step": 58360 }, { "epoch": 1.166420207026098, "grad_norm": 0.11740583926439285, "learning_rate": 7.526592696981043e-05, "loss": 0.0261, "step": 58370 }, { "epoch": 1.1666200391670996, "grad_norm": 0.13315176963806152, "learning_rate": 7.525700872090848e-05, "loss": 0.032, "step": 58380 }, { "epoch": 1.1668198713081013, "grad_norm": 0.12149186432361603, "learning_rate": 7.524808939307353e-05, "loss": 0.0426, "step": 58390 }, { "epoch": 1.1670197034491028, "grad_norm": 0.12050779163837433, "learning_rate": 7.523916898668657e-05, "loss": 0.0343, "step": 58400 }, { "epoch": 1.1672195355901043, "grad_norm": 0.11653860658407211, "learning_rate": 7.523024750212869e-05, "loss": 0.0356, "step": 58410 }, { "epoch": 1.1674193677311058, "grad_norm": 0.12177076190710068, "learning_rate": 7.522132493978098e-05, "loss": 0.0307, "step": 58420 }, { "epoch": 1.1676191998721075, "grad_norm": 0.1744401603937149, "learning_rate": 7.521240130002461e-05, "loss": 0.0365, "step": 58430 }, { "epoch": 1.167819032013109, "grad_norm": 0.12214953452348709, "learning_rate": 7.520347658324076e-05, "loss": 0.0386, "step": 58440 }, { "epoch": 1.1680188641541105, "grad_norm": 0.13761073350906372, "learning_rate": 7.519455078981072e-05, "loss": 0.0349, "step": 58450 }, { "epoch": 1.1682186962951122, "grad_norm": 0.12874089181423187, "learning_rate": 7.518562392011572e-05, "loss": 0.0293, "step": 58460 }, { "epoch": 1.1684185284361137, "grad_norm": 0.11568215489387512, "learning_rate": 7.517669597453714e-05, "loss": 0.0332, "step": 58470 }, { "epoch": 1.1686183605771152, "grad_norm": 0.1136455088853836, "learning_rate": 7.516776695345637e-05, "loss": 0.0305, "step": 58480 }, { "epoch": 1.1688181927181167, "grad_norm": 0.12120538204908371, "learning_rate": 7.515883685725483e-05, "loss": 0.029, "step": 58490 }, { "epoch": 1.1690180248591184, "grad_norm": 0.1250402182340622, "learning_rate": 7.5149905686314e-05, "loss": 0.0296, "step": 58500 }, { "epoch": 1.16921785700012, "grad_norm": 0.12294623255729675, "learning_rate": 7.514097344101538e-05, "loss": 0.0362, "step": 58510 }, { "epoch": 1.1694176891411214, "grad_norm": 0.16984497010707855, "learning_rate": 7.513204012174059e-05, "loss": 0.0351, "step": 58520 }, { "epoch": 1.169617521282123, "grad_norm": 0.10096555948257446, "learning_rate": 7.512310572887121e-05, "loss": 0.0432, "step": 58530 }, { "epoch": 1.1698173534231247, "grad_norm": 0.18169741332530975, "learning_rate": 7.511417026278888e-05, "loss": 0.0356, "step": 58540 }, { "epoch": 1.1700171855641261, "grad_norm": 0.14874513447284698, "learning_rate": 7.510523372387536e-05, "loss": 0.0336, "step": 58550 }, { "epoch": 1.1702170177051276, "grad_norm": 0.14536935091018677, "learning_rate": 7.509629611251235e-05, "loss": 0.0282, "step": 58560 }, { "epoch": 1.1704168498461294, "grad_norm": 0.09974068403244019, "learning_rate": 7.508735742908169e-05, "loss": 0.0346, "step": 58570 }, { "epoch": 1.1706166819871309, "grad_norm": 0.12276051193475723, "learning_rate": 7.50784176739652e-05, "loss": 0.0309, "step": 58580 }, { "epoch": 1.1708165141281324, "grad_norm": 0.13761889934539795, "learning_rate": 7.506947684754479e-05, "loss": 0.0309, "step": 58590 }, { "epoch": 1.1710163462691339, "grad_norm": 0.137711763381958, "learning_rate": 7.506053495020237e-05, "loss": 0.0279, "step": 58600 }, { "epoch": 1.1712161784101354, "grad_norm": 0.14043663442134857, "learning_rate": 7.505159198231993e-05, "loss": 0.0364, "step": 58610 }, { "epoch": 1.171416010551137, "grad_norm": 0.1313106268644333, "learning_rate": 7.50426479442795e-05, "loss": 0.0295, "step": 58620 }, { "epoch": 1.1716158426921386, "grad_norm": 0.1420716941356659, "learning_rate": 7.503370283646318e-05, "loss": 0.0376, "step": 58630 }, { "epoch": 1.17181567483314, "grad_norm": 0.1101788878440857, "learning_rate": 7.502475665925305e-05, "loss": 0.0268, "step": 58640 }, { "epoch": 1.1720155069741418, "grad_norm": 0.14261801540851593, "learning_rate": 7.50158094130313e-05, "loss": 0.0457, "step": 58650 }, { "epoch": 1.1722153391151433, "grad_norm": 0.1420888602733612, "learning_rate": 7.50068610981801e-05, "loss": 0.0352, "step": 58660 }, { "epoch": 1.1724151712561448, "grad_norm": 0.11555933952331543, "learning_rate": 7.499791171508175e-05, "loss": 0.0313, "step": 58670 }, { "epoch": 1.1726150033971463, "grad_norm": 0.1240481585264206, "learning_rate": 7.498896126411852e-05, "loss": 0.0324, "step": 58680 }, { "epoch": 1.172814835538148, "grad_norm": 0.08586165308952332, "learning_rate": 7.498000974567279e-05, "loss": 0.0315, "step": 58690 }, { "epoch": 1.1730146676791495, "grad_norm": 0.12723790109157562, "learning_rate": 7.497105716012693e-05, "loss": 0.0388, "step": 58700 }, { "epoch": 1.173214499820151, "grad_norm": 0.12324398756027222, "learning_rate": 7.496210350786336e-05, "loss": 0.031, "step": 58710 }, { "epoch": 1.1734143319611525, "grad_norm": 0.11722321808338165, "learning_rate": 7.495314878926463e-05, "loss": 0.0295, "step": 58720 }, { "epoch": 1.1736141641021542, "grad_norm": 0.16255447268486023, "learning_rate": 7.494419300471318e-05, "loss": 0.0251, "step": 58730 }, { "epoch": 1.1738139962431557, "grad_norm": 0.11851517856121063, "learning_rate": 7.493523615459165e-05, "loss": 0.0282, "step": 58740 }, { "epoch": 1.1740138283841572, "grad_norm": 0.12715503573417664, "learning_rate": 7.492627823928264e-05, "loss": 0.0327, "step": 58750 }, { "epoch": 1.174213660525159, "grad_norm": 0.13757435977458954, "learning_rate": 7.49173192591688e-05, "loss": 0.0344, "step": 58760 }, { "epoch": 1.1744134926661605, "grad_norm": 0.12600520253181458, "learning_rate": 7.490835921463287e-05, "loss": 0.0368, "step": 58770 }, { "epoch": 1.174613324807162, "grad_norm": 0.11802178621292114, "learning_rate": 7.489939810605758e-05, "loss": 0.0334, "step": 58780 }, { "epoch": 1.1748131569481635, "grad_norm": 0.11519770324230194, "learning_rate": 7.489043593382575e-05, "loss": 0.0412, "step": 58790 }, { "epoch": 1.1750129890891652, "grad_norm": 0.1280679702758789, "learning_rate": 7.488147269832023e-05, "loss": 0.0363, "step": 58800 }, { "epoch": 1.1752128212301667, "grad_norm": 0.13823409378528595, "learning_rate": 7.48725083999239e-05, "loss": 0.0302, "step": 58810 }, { "epoch": 1.1754126533711682, "grad_norm": 0.11054901778697968, "learning_rate": 7.486354303901971e-05, "loss": 0.0321, "step": 58820 }, { "epoch": 1.1756124855121697, "grad_norm": 0.12722378969192505, "learning_rate": 7.485457661599062e-05, "loss": 0.0317, "step": 58830 }, { "epoch": 1.1758123176531714, "grad_norm": 0.11459024995565414, "learning_rate": 7.484560913121969e-05, "loss": 0.0338, "step": 58840 }, { "epoch": 1.176012149794173, "grad_norm": 0.13198336958885193, "learning_rate": 7.483664058508996e-05, "loss": 0.034, "step": 58850 }, { "epoch": 1.1762119819351744, "grad_norm": 0.16562339663505554, "learning_rate": 7.482767097798459e-05, "loss": 0.0364, "step": 58860 }, { "epoch": 1.1764118140761761, "grad_norm": 0.16425520181655884, "learning_rate": 7.481870031028673e-05, "loss": 0.033, "step": 58870 }, { "epoch": 1.1766116462171776, "grad_norm": 0.11880286782979965, "learning_rate": 7.480972858237958e-05, "loss": 0.0277, "step": 58880 }, { "epoch": 1.1768114783581791, "grad_norm": 0.15175633132457733, "learning_rate": 7.480075579464639e-05, "loss": 0.03, "step": 58890 }, { "epoch": 1.1770113104991806, "grad_norm": 0.11922235786914825, "learning_rate": 7.47917819474705e-05, "loss": 0.0368, "step": 58900 }, { "epoch": 1.1772111426401823, "grad_norm": 0.10485587269067764, "learning_rate": 7.47828070412352e-05, "loss": 0.0393, "step": 58910 }, { "epoch": 1.1774109747811838, "grad_norm": 0.12269625812768936, "learning_rate": 7.477383107632393e-05, "loss": 0.0367, "step": 58920 }, { "epoch": 1.1776108069221853, "grad_norm": 0.13346537947654724, "learning_rate": 7.47648540531201e-05, "loss": 0.0322, "step": 58930 }, { "epoch": 1.1778106390631868, "grad_norm": 0.11032632738351822, "learning_rate": 7.47558759720072e-05, "loss": 0.0262, "step": 58940 }, { "epoch": 1.1780104712041886, "grad_norm": 0.10402192920446396, "learning_rate": 7.474689683336876e-05, "loss": 0.0342, "step": 58950 }, { "epoch": 1.17821030334519, "grad_norm": 0.11262963712215424, "learning_rate": 7.473791663758833e-05, "loss": 0.0297, "step": 58960 }, { "epoch": 1.1784101354861916, "grad_norm": 0.10586252063512802, "learning_rate": 7.472893538504957e-05, "loss": 0.0314, "step": 58970 }, { "epoch": 1.1786099676271933, "grad_norm": 0.15265654027462006, "learning_rate": 7.47199530761361e-05, "loss": 0.0345, "step": 58980 }, { "epoch": 1.1788097997681948, "grad_norm": 0.1376427412033081, "learning_rate": 7.471096971123168e-05, "loss": 0.0308, "step": 58990 }, { "epoch": 1.1790096319091963, "grad_norm": 0.15944208204746246, "learning_rate": 7.470198529071999e-05, "loss": 0.0281, "step": 59000 }, { "epoch": 1.1792094640501978, "grad_norm": 0.1287517547607422, "learning_rate": 7.469299981498488e-05, "loss": 0.0352, "step": 59010 }, { "epoch": 1.1794092961911993, "grad_norm": 0.15565519034862518, "learning_rate": 7.46840132844102e-05, "loss": 0.0368, "step": 59020 }, { "epoch": 1.179609128332201, "grad_norm": 0.1495206207036972, "learning_rate": 7.467502569937978e-05, "loss": 0.0339, "step": 59030 }, { "epoch": 1.1798089604732025, "grad_norm": 0.13888698816299438, "learning_rate": 7.466603706027764e-05, "loss": 0.0375, "step": 59040 }, { "epoch": 1.180008792614204, "grad_norm": 0.13281552493572235, "learning_rate": 7.465704736748767e-05, "loss": 0.0393, "step": 59050 }, { "epoch": 1.1802086247552057, "grad_norm": 0.13347159326076508, "learning_rate": 7.464805662139394e-05, "loss": 0.0291, "step": 59060 }, { "epoch": 1.1804084568962072, "grad_norm": 0.1460115909576416, "learning_rate": 7.463906482238052e-05, "loss": 0.0334, "step": 59070 }, { "epoch": 1.1806082890372087, "grad_norm": 0.10106506943702698, "learning_rate": 7.463007197083151e-05, "loss": 0.0288, "step": 59080 }, { "epoch": 1.1808081211782102, "grad_norm": 0.09218666702508926, "learning_rate": 7.462107806713108e-05, "loss": 0.031, "step": 59090 }, { "epoch": 1.181007953319212, "grad_norm": 0.12013483047485352, "learning_rate": 7.461208311166344e-05, "loss": 0.0278, "step": 59100 }, { "epoch": 1.1812077854602134, "grad_norm": 0.15888039767742157, "learning_rate": 7.460308710481279e-05, "loss": 0.0421, "step": 59110 }, { "epoch": 1.181407617601215, "grad_norm": 0.12090729922056198, "learning_rate": 7.459409004696347e-05, "loss": 0.0334, "step": 59120 }, { "epoch": 1.1816074497422164, "grad_norm": 0.13272036612033844, "learning_rate": 7.45850919384998e-05, "loss": 0.0339, "step": 59130 }, { "epoch": 1.1818072818832182, "grad_norm": 0.13869084417819977, "learning_rate": 7.457609277980618e-05, "loss": 0.0272, "step": 59140 }, { "epoch": 1.1820071140242197, "grad_norm": 0.1421540081501007, "learning_rate": 7.4567092571267e-05, "loss": 0.0329, "step": 59150 }, { "epoch": 1.1822069461652212, "grad_norm": 0.1467040330171585, "learning_rate": 7.455809131326676e-05, "loss": 0.0356, "step": 59160 }, { "epoch": 1.1824067783062229, "grad_norm": 0.13659659028053284, "learning_rate": 7.454908900619e-05, "loss": 0.037, "step": 59170 }, { "epoch": 1.1826066104472244, "grad_norm": 0.13745303452014923, "learning_rate": 7.454008565042123e-05, "loss": 0.0269, "step": 59180 }, { "epoch": 1.1828064425882259, "grad_norm": 0.15442928671836853, "learning_rate": 7.45310812463451e-05, "loss": 0.0369, "step": 59190 }, { "epoch": 1.1830062747292274, "grad_norm": 0.13482755422592163, "learning_rate": 7.452207579434624e-05, "loss": 0.0346, "step": 59200 }, { "epoch": 1.183206106870229, "grad_norm": 0.12724696099758148, "learning_rate": 7.451306929480935e-05, "loss": 0.0361, "step": 59210 }, { "epoch": 1.1834059390112306, "grad_norm": 0.1356932669878006, "learning_rate": 7.450406174811917e-05, "loss": 0.0342, "step": 59220 }, { "epoch": 1.183605771152232, "grad_norm": 0.12380822747945786, "learning_rate": 7.449505315466048e-05, "loss": 0.0339, "step": 59230 }, { "epoch": 1.1838056032932336, "grad_norm": 0.10392356663942337, "learning_rate": 7.448604351481812e-05, "loss": 0.03, "step": 59240 }, { "epoch": 1.1840054354342353, "grad_norm": 0.09568382799625397, "learning_rate": 7.447703282897698e-05, "loss": 0.0332, "step": 59250 }, { "epoch": 1.1842052675752368, "grad_norm": 0.09297681599855423, "learning_rate": 7.446802109752194e-05, "loss": 0.0259, "step": 59260 }, { "epoch": 1.1844050997162383, "grad_norm": 0.13736961781978607, "learning_rate": 7.445900832083801e-05, "loss": 0.032, "step": 59270 }, { "epoch": 1.18460493185724, "grad_norm": 0.11114909499883652, "learning_rate": 7.444999449931016e-05, "loss": 0.0327, "step": 59280 }, { "epoch": 1.1848047639982415, "grad_norm": 0.1413520872592926, "learning_rate": 7.444097963332346e-05, "loss": 0.0403, "step": 59290 }, { "epoch": 1.185004596139243, "grad_norm": 0.1574193239212036, "learning_rate": 7.443196372326304e-05, "loss": 0.0316, "step": 59300 }, { "epoch": 1.1852044282802445, "grad_norm": 0.13691100478172302, "learning_rate": 7.442294676951398e-05, "loss": 0.0443, "step": 59310 }, { "epoch": 1.1854042604212462, "grad_norm": 0.2111845314502716, "learning_rate": 7.441392877246151e-05, "loss": 0.0298, "step": 59320 }, { "epoch": 1.1856040925622477, "grad_norm": 0.15445280075073242, "learning_rate": 7.440490973249085e-05, "loss": 0.0353, "step": 59330 }, { "epoch": 1.1858039247032492, "grad_norm": 0.15008024871349335, "learning_rate": 7.439588964998729e-05, "loss": 0.0321, "step": 59340 }, { "epoch": 1.1860037568442507, "grad_norm": 0.1560283601284027, "learning_rate": 7.438686852533611e-05, "loss": 0.0361, "step": 59350 }, { "epoch": 1.1862035889852525, "grad_norm": 0.15506741404533386, "learning_rate": 7.437784635892275e-05, "loss": 0.0357, "step": 59360 }, { "epoch": 1.186403421126254, "grad_norm": 0.10603012889623642, "learning_rate": 7.436882315113256e-05, "loss": 0.0315, "step": 59370 }, { "epoch": 1.1866032532672555, "grad_norm": 0.10694530606269836, "learning_rate": 7.435979890235098e-05, "loss": 0.0317, "step": 59380 }, { "epoch": 1.1868030854082572, "grad_norm": 0.16080285608768463, "learning_rate": 7.435077361296358e-05, "loss": 0.0279, "step": 59390 }, { "epoch": 1.1870029175492587, "grad_norm": 0.12109621614217758, "learning_rate": 7.434174728335587e-05, "loss": 0.0275, "step": 59400 }, { "epoch": 1.1872027496902602, "grad_norm": 0.12259069085121155, "learning_rate": 7.43327199139134e-05, "loss": 0.0369, "step": 59410 }, { "epoch": 1.1874025818312617, "grad_norm": 0.1507607251405716, "learning_rate": 7.432369150502188e-05, "loss": 0.0329, "step": 59420 }, { "epoch": 1.1876024139722632, "grad_norm": 0.09709739685058594, "learning_rate": 7.431466205706689e-05, "loss": 0.0333, "step": 59430 }, { "epoch": 1.187802246113265, "grad_norm": 0.10943092405796051, "learning_rate": 7.430563157043426e-05, "loss": 0.0327, "step": 59440 }, { "epoch": 1.1880020782542664, "grad_norm": 0.1279440075159073, "learning_rate": 7.429660004550965e-05, "loss": 0.0331, "step": 59450 }, { "epoch": 1.188201910395268, "grad_norm": 0.127807155251503, "learning_rate": 7.428756748267896e-05, "loss": 0.0338, "step": 59460 }, { "epoch": 1.1884017425362696, "grad_norm": 0.12910892069339752, "learning_rate": 7.427853388232798e-05, "loss": 0.0327, "step": 59470 }, { "epoch": 1.1886015746772711, "grad_norm": 0.15191291272640228, "learning_rate": 7.426949924484263e-05, "loss": 0.04, "step": 59480 }, { "epoch": 1.1888014068182726, "grad_norm": 0.13991443812847137, "learning_rate": 7.426046357060889e-05, "loss": 0.0397, "step": 59490 }, { "epoch": 1.1890012389592741, "grad_norm": 0.14021362364292145, "learning_rate": 7.425142686001269e-05, "loss": 0.03, "step": 59500 }, { "epoch": 1.1892010711002758, "grad_norm": 0.11799737066030502, "learning_rate": 7.424238911344008e-05, "loss": 0.0421, "step": 59510 }, { "epoch": 1.1894009032412773, "grad_norm": 0.17442956566810608, "learning_rate": 7.423335033127714e-05, "loss": 0.0308, "step": 59520 }, { "epoch": 1.1896007353822788, "grad_norm": 0.12633906304836273, "learning_rate": 7.422431051391001e-05, "loss": 0.0314, "step": 59530 }, { "epoch": 1.1898005675232803, "grad_norm": 0.14826112985610962, "learning_rate": 7.421526966172481e-05, "loss": 0.029, "step": 59540 }, { "epoch": 1.190000399664282, "grad_norm": 0.1393279731273651, "learning_rate": 7.420622777510779e-05, "loss": 0.0284, "step": 59550 }, { "epoch": 1.1902002318052836, "grad_norm": 0.15447838604450226, "learning_rate": 7.41971848544452e-05, "loss": 0.038, "step": 59560 }, { "epoch": 1.190400063946285, "grad_norm": 0.121463842689991, "learning_rate": 7.418814090012333e-05, "loss": 0.0385, "step": 59570 }, { "epoch": 1.1905998960872868, "grad_norm": 0.12396159768104553, "learning_rate": 7.41790959125285e-05, "loss": 0.0333, "step": 59580 }, { "epoch": 1.1907997282282883, "grad_norm": 0.147503063082695, "learning_rate": 7.41700498920471e-05, "loss": 0.0403, "step": 59590 }, { "epoch": 1.1909995603692898, "grad_norm": 0.1068497970700264, "learning_rate": 7.416100283906558e-05, "loss": 0.0364, "step": 59600 }, { "epoch": 1.1911993925102913, "grad_norm": 0.09112408012151718, "learning_rate": 7.415195475397041e-05, "loss": 0.0363, "step": 59610 }, { "epoch": 1.191399224651293, "grad_norm": 0.12185519933700562, "learning_rate": 7.41429056371481e-05, "loss": 0.0369, "step": 59620 }, { "epoch": 1.1915990567922945, "grad_norm": 0.1252482682466507, "learning_rate": 7.413385548898521e-05, "loss": 0.0364, "step": 59630 }, { "epoch": 1.191798888933296, "grad_norm": 0.1257171332836151, "learning_rate": 7.412480430986835e-05, "loss": 0.0348, "step": 59640 }, { "epoch": 1.1919987210742975, "grad_norm": 0.11035805940628052, "learning_rate": 7.411575210018417e-05, "loss": 0.0254, "step": 59650 }, { "epoch": 1.1921985532152992, "grad_norm": 0.12455751746892929, "learning_rate": 7.410669886031936e-05, "loss": 0.0346, "step": 59660 }, { "epoch": 1.1923983853563007, "grad_norm": 0.18033495545387268, "learning_rate": 7.409764459066067e-05, "loss": 0.0333, "step": 59670 }, { "epoch": 1.1925982174973022, "grad_norm": 0.18792074918746948, "learning_rate": 7.408858929159485e-05, "loss": 0.0361, "step": 59680 }, { "epoch": 1.192798049638304, "grad_norm": 0.18385963141918182, "learning_rate": 7.407953296350875e-05, "loss": 0.0357, "step": 59690 }, { "epoch": 1.1929978817793054, "grad_norm": 0.13687627017498016, "learning_rate": 7.407047560678926e-05, "loss": 0.0355, "step": 59700 }, { "epoch": 1.193197713920307, "grad_norm": 0.1380939930677414, "learning_rate": 7.406141722182325e-05, "loss": 0.0325, "step": 59710 }, { "epoch": 1.1933975460613084, "grad_norm": 0.15272954106330872, "learning_rate": 7.405235780899771e-05, "loss": 0.0439, "step": 59720 }, { "epoch": 1.1935973782023102, "grad_norm": 0.151652991771698, "learning_rate": 7.404329736869963e-05, "loss": 0.0372, "step": 59730 }, { "epoch": 1.1937972103433117, "grad_norm": 0.10113757848739624, "learning_rate": 7.403423590131604e-05, "loss": 0.0381, "step": 59740 }, { "epoch": 1.1939970424843132, "grad_norm": 0.13105140626430511, "learning_rate": 7.402517340723406e-05, "loss": 0.0364, "step": 59750 }, { "epoch": 1.1941968746253147, "grad_norm": 0.13576100766658783, "learning_rate": 7.401610988684081e-05, "loss": 0.0255, "step": 59760 }, { "epoch": 1.1943967067663164, "grad_norm": 0.14140671491622925, "learning_rate": 7.400704534052346e-05, "loss": 0.0296, "step": 59770 }, { "epoch": 1.1945965389073179, "grad_norm": 0.10227233171463013, "learning_rate": 7.399797976866923e-05, "loss": 0.0295, "step": 59780 }, { "epoch": 1.1947963710483194, "grad_norm": 0.11208165436983109, "learning_rate": 7.39889131716654e-05, "loss": 0.0342, "step": 59790 }, { "epoch": 1.194996203189321, "grad_norm": 0.12839128077030182, "learning_rate": 7.397984554989925e-05, "loss": 0.0381, "step": 59800 }, { "epoch": 1.1951960353303226, "grad_norm": 0.12282127887010574, "learning_rate": 7.397077690375816e-05, "loss": 0.0265, "step": 59810 }, { "epoch": 1.195395867471324, "grad_norm": 0.138123020529747, "learning_rate": 7.396170723362954e-05, "loss": 0.0352, "step": 59820 }, { "epoch": 1.1955956996123256, "grad_norm": 0.12415871024131775, "learning_rate": 7.395263653990077e-05, "loss": 0.0269, "step": 59830 }, { "epoch": 1.195795531753327, "grad_norm": 0.12663079798221588, "learning_rate": 7.394356482295939e-05, "loss": 0.0307, "step": 59840 }, { "epoch": 1.1959953638943288, "grad_norm": 0.17162984609603882, "learning_rate": 7.39344920831929e-05, "loss": 0.0292, "step": 59850 }, { "epoch": 1.1961951960353303, "grad_norm": 0.12275867909193039, "learning_rate": 7.392541832098889e-05, "loss": 0.0411, "step": 59860 }, { "epoch": 1.1963950281763318, "grad_norm": 0.11433808505535126, "learning_rate": 7.391634353673495e-05, "loss": 0.0362, "step": 59870 }, { "epoch": 1.1965948603173335, "grad_norm": 0.12619023025035858, "learning_rate": 7.390726773081874e-05, "loss": 0.0334, "step": 59880 }, { "epoch": 1.196794692458335, "grad_norm": 0.13757286965847015, "learning_rate": 7.389819090362799e-05, "loss": 0.0367, "step": 59890 }, { "epoch": 1.1969945245993365, "grad_norm": 0.15793433785438538, "learning_rate": 7.388911305555042e-05, "loss": 0.0305, "step": 59900 }, { "epoch": 1.197194356740338, "grad_norm": 0.14272582530975342, "learning_rate": 7.388003418697383e-05, "loss": 0.0451, "step": 59910 }, { "epoch": 1.1973941888813397, "grad_norm": 0.12887334823608398, "learning_rate": 7.387095429828603e-05, "loss": 0.0292, "step": 59920 }, { "epoch": 1.1975940210223412, "grad_norm": 0.11827856302261353, "learning_rate": 7.386187338987492e-05, "loss": 0.0338, "step": 59930 }, { "epoch": 1.1977938531633427, "grad_norm": 0.13411468267440796, "learning_rate": 7.385279146212842e-05, "loss": 0.0328, "step": 59940 }, { "epoch": 1.1979936853043442, "grad_norm": 0.14534440636634827, "learning_rate": 7.384370851543449e-05, "loss": 0.0339, "step": 59950 }, { "epoch": 1.198193517445346, "grad_norm": 0.13402169942855835, "learning_rate": 7.383462455018116e-05, "loss": 0.0339, "step": 59960 }, { "epoch": 1.1983933495863475, "grad_norm": 0.137654647231102, "learning_rate": 7.382553956675642e-05, "loss": 0.0281, "step": 59970 }, { "epoch": 1.198593181727349, "grad_norm": 0.11463190615177155, "learning_rate": 7.38164535655484e-05, "loss": 0.0335, "step": 59980 }, { "epoch": 1.1987930138683507, "grad_norm": 0.15944992005825043, "learning_rate": 7.380736654694524e-05, "loss": 0.0334, "step": 59990 }, { "epoch": 1.1989928460093522, "grad_norm": 0.15124624967575073, "learning_rate": 7.379827851133513e-05, "loss": 0.0361, "step": 60000 }, { "epoch": 1.1991926781503537, "grad_norm": 0.09218436479568481, "learning_rate": 7.378918945910626e-05, "loss": 0.0401, "step": 60010 }, { "epoch": 1.1993925102913552, "grad_norm": 0.11541248112916946, "learning_rate": 7.378009939064695e-05, "loss": 0.0252, "step": 60020 }, { "epoch": 1.199592342432357, "grad_norm": 0.12893500924110413, "learning_rate": 7.377100830634545e-05, "loss": 0.0395, "step": 60030 }, { "epoch": 1.1997921745733584, "grad_norm": 0.16576170921325684, "learning_rate": 7.376191620659016e-05, "loss": 0.0325, "step": 60040 }, { "epoch": 1.19999200671436, "grad_norm": 0.12113723158836365, "learning_rate": 7.375282309176945e-05, "loss": 0.0395, "step": 60050 }, { "epoch": 1.2001918388553614, "grad_norm": 0.1500486433506012, "learning_rate": 7.374372896227177e-05, "loss": 0.0327, "step": 60060 }, { "epoch": 1.2003916709963631, "grad_norm": 0.1345449835062027, "learning_rate": 7.373463381848563e-05, "loss": 0.0328, "step": 60070 }, { "epoch": 1.2005915031373646, "grad_norm": 0.12668485939502716, "learning_rate": 7.37255376607995e-05, "loss": 0.04, "step": 60080 }, { "epoch": 1.2007913352783661, "grad_norm": 0.1474379450082779, "learning_rate": 7.371644048960203e-05, "loss": 0.0297, "step": 60090 }, { "epoch": 1.2009911674193678, "grad_norm": 0.12403276562690735, "learning_rate": 7.370734230528175e-05, "loss": 0.0315, "step": 60100 }, { "epoch": 1.2011909995603693, "grad_norm": 0.10570766031742096, "learning_rate": 7.369824310822738e-05, "loss": 0.028, "step": 60110 }, { "epoch": 1.2013908317013708, "grad_norm": 0.1293945461511612, "learning_rate": 7.36891428988276e-05, "loss": 0.0376, "step": 60120 }, { "epoch": 1.2015906638423723, "grad_norm": 0.15224120020866394, "learning_rate": 7.368004167747114e-05, "loss": 0.0341, "step": 60130 }, { "epoch": 1.2017904959833738, "grad_norm": 0.1029960960149765, "learning_rate": 7.367093944454683e-05, "loss": 0.0305, "step": 60140 }, { "epoch": 1.2019903281243756, "grad_norm": 0.12138557434082031, "learning_rate": 7.366183620044344e-05, "loss": 0.0279, "step": 60150 }, { "epoch": 1.202190160265377, "grad_norm": 0.1409372240304947, "learning_rate": 7.36527319455499e-05, "loss": 0.0314, "step": 60160 }, { "epoch": 1.2023899924063786, "grad_norm": 0.1347253918647766, "learning_rate": 7.364362668025508e-05, "loss": 0.0359, "step": 60170 }, { "epoch": 1.2025898245473803, "grad_norm": 0.1082233116030693, "learning_rate": 7.363452040494798e-05, "loss": 0.0306, "step": 60180 }, { "epoch": 1.2027896566883818, "grad_norm": 0.17807777225971222, "learning_rate": 7.362541312001757e-05, "loss": 0.0366, "step": 60190 }, { "epoch": 1.2029894888293833, "grad_norm": 0.1464313417673111, "learning_rate": 7.361630482585292e-05, "loss": 0.0358, "step": 60200 }, { "epoch": 1.203189320970385, "grad_norm": 0.14025847613811493, "learning_rate": 7.360719552284312e-05, "loss": 0.0389, "step": 60210 }, { "epoch": 1.2033891531113865, "grad_norm": 0.14598268270492554, "learning_rate": 7.35980852113773e-05, "loss": 0.0292, "step": 60220 }, { "epoch": 1.203588985252388, "grad_norm": 0.1597180813550949, "learning_rate": 7.358897389184461e-05, "loss": 0.0362, "step": 60230 }, { "epoch": 1.2037888173933895, "grad_norm": 0.16819852590560913, "learning_rate": 7.35798615646343e-05, "loss": 0.0294, "step": 60240 }, { "epoch": 1.203988649534391, "grad_norm": 0.18378570675849915, "learning_rate": 7.357074823013563e-05, "loss": 0.0409, "step": 60250 }, { "epoch": 1.2041884816753927, "grad_norm": 0.15675124526023865, "learning_rate": 7.35616338887379e-05, "loss": 0.035, "step": 60260 }, { "epoch": 1.2043883138163942, "grad_norm": 0.17451897263526917, "learning_rate": 7.355251854083044e-05, "loss": 0.0324, "step": 60270 }, { "epoch": 1.2045881459573957, "grad_norm": 0.18455933034420013, "learning_rate": 7.354340218680264e-05, "loss": 0.0401, "step": 60280 }, { "epoch": 1.2047879780983974, "grad_norm": 0.1281159371137619, "learning_rate": 7.353428482704398e-05, "loss": 0.0365, "step": 60290 }, { "epoch": 1.204987810239399, "grad_norm": 0.1607332080602646, "learning_rate": 7.352516646194389e-05, "loss": 0.0426, "step": 60300 }, { "epoch": 1.2051876423804004, "grad_norm": 0.11858464032411575, "learning_rate": 7.35160470918919e-05, "loss": 0.0354, "step": 60310 }, { "epoch": 1.205387474521402, "grad_norm": 0.12493281066417694, "learning_rate": 7.350692671727761e-05, "loss": 0.038, "step": 60320 }, { "epoch": 1.2055873066624037, "grad_norm": 0.17935523390769958, "learning_rate": 7.349780533849056e-05, "loss": 0.041, "step": 60330 }, { "epoch": 1.2057871388034052, "grad_norm": 0.15087051689624786, "learning_rate": 7.348868295592045e-05, "loss": 0.0344, "step": 60340 }, { "epoch": 1.2059869709444067, "grad_norm": 0.15229249000549316, "learning_rate": 7.347955956995693e-05, "loss": 0.0354, "step": 60350 }, { "epoch": 1.2061868030854082, "grad_norm": 0.15675143897533417, "learning_rate": 7.34704351809898e-05, "loss": 0.041, "step": 60360 }, { "epoch": 1.2063866352264099, "grad_norm": 0.14111344516277313, "learning_rate": 7.346130978940877e-05, "loss": 0.0359, "step": 60370 }, { "epoch": 1.2065864673674114, "grad_norm": 0.15435534715652466, "learning_rate": 7.345218339560368e-05, "loss": 0.0352, "step": 60380 }, { "epoch": 1.2067862995084129, "grad_norm": 0.14763784408569336, "learning_rate": 7.344305599996441e-05, "loss": 0.0362, "step": 60390 }, { "epoch": 1.2069861316494146, "grad_norm": 0.12681859731674194, "learning_rate": 7.343392760288085e-05, "loss": 0.0312, "step": 60400 }, { "epoch": 1.207185963790416, "grad_norm": 0.12108460068702698, "learning_rate": 7.342479820474296e-05, "loss": 0.0308, "step": 60410 }, { "epoch": 1.2073857959314176, "grad_norm": 0.15126368403434753, "learning_rate": 7.341566780594071e-05, "loss": 0.0356, "step": 60420 }, { "epoch": 1.207585628072419, "grad_norm": 0.14720191061496735, "learning_rate": 7.340653640686415e-05, "loss": 0.0282, "step": 60430 }, { "epoch": 1.2077854602134208, "grad_norm": 0.13110017776489258, "learning_rate": 7.339740400790337e-05, "loss": 0.0406, "step": 60440 }, { "epoch": 1.2079852923544223, "grad_norm": 0.13973022997379303, "learning_rate": 7.338827060944846e-05, "loss": 0.0407, "step": 60450 }, { "epoch": 1.2081851244954238, "grad_norm": 0.12807974219322205, "learning_rate": 7.337913621188959e-05, "loss": 0.0297, "step": 60460 }, { "epoch": 1.2083849566364253, "grad_norm": 0.12784145772457123, "learning_rate": 7.3370000815617e-05, "loss": 0.0296, "step": 60470 }, { "epoch": 1.208584788777427, "grad_norm": 0.11716312170028687, "learning_rate": 7.336086442102086e-05, "loss": 0.0327, "step": 60480 }, { "epoch": 1.2087846209184285, "grad_norm": 0.1382952630519867, "learning_rate": 7.335172702849154e-05, "loss": 0.0233, "step": 60490 }, { "epoch": 1.20898445305943, "grad_norm": 0.1328677088022232, "learning_rate": 7.334258863841933e-05, "loss": 0.0306, "step": 60500 }, { "epoch": 1.2091842852004318, "grad_norm": 0.17808844149112701, "learning_rate": 7.33334492511946e-05, "loss": 0.0369, "step": 60510 }, { "epoch": 1.2093841173414333, "grad_norm": 0.12978112697601318, "learning_rate": 7.332430886720783e-05, "loss": 0.0285, "step": 60520 }, { "epoch": 1.2095839494824348, "grad_norm": 0.14892178773880005, "learning_rate": 7.331516748684938e-05, "loss": 0.0389, "step": 60530 }, { "epoch": 1.2097837816234362, "grad_norm": 0.13457687199115753, "learning_rate": 7.330602511050984e-05, "loss": 0.0289, "step": 60540 }, { "epoch": 1.2099836137644377, "grad_norm": 0.15503409504890442, "learning_rate": 7.329688173857969e-05, "loss": 0.0379, "step": 60550 }, { "epoch": 1.2101834459054395, "grad_norm": 0.15087242424488068, "learning_rate": 7.328773737144959e-05, "loss": 0.0356, "step": 60560 }, { "epoch": 1.210383278046441, "grad_norm": 0.12702424824237823, "learning_rate": 7.327859200951011e-05, "loss": 0.032, "step": 60570 }, { "epoch": 1.2105831101874425, "grad_norm": 0.15120269358158112, "learning_rate": 7.326944565315193e-05, "loss": 0.0314, "step": 60580 }, { "epoch": 1.2107829423284442, "grad_norm": 0.10658930242061615, "learning_rate": 7.326029830276579e-05, "loss": 0.0332, "step": 60590 }, { "epoch": 1.2109827744694457, "grad_norm": 0.13192810118198395, "learning_rate": 7.325114995874245e-05, "loss": 0.0283, "step": 60600 }, { "epoch": 1.2111826066104472, "grad_norm": 0.10537884384393692, "learning_rate": 7.32420006214727e-05, "loss": 0.0405, "step": 60610 }, { "epoch": 1.2113824387514487, "grad_norm": 0.18219289183616638, "learning_rate": 7.323285029134736e-05, "loss": 0.0237, "step": 60620 }, { "epoch": 1.2115822708924504, "grad_norm": 0.21433840692043304, "learning_rate": 7.322369896875733e-05, "loss": 0.0308, "step": 60630 }, { "epoch": 1.211782103033452, "grad_norm": 0.1293594241142273, "learning_rate": 7.321454665409357e-05, "loss": 0.0291, "step": 60640 }, { "epoch": 1.2119819351744534, "grad_norm": 0.12060786783695221, "learning_rate": 7.320539334774701e-05, "loss": 0.0337, "step": 60650 }, { "epoch": 1.212181767315455, "grad_norm": 0.13357552886009216, "learning_rate": 7.319623905010867e-05, "loss": 0.0454, "step": 60660 }, { "epoch": 1.2123815994564566, "grad_norm": 0.12451349943876266, "learning_rate": 7.318708376156962e-05, "loss": 0.0315, "step": 60670 }, { "epoch": 1.2125814315974581, "grad_norm": 0.14398710429668427, "learning_rate": 7.317792748252094e-05, "loss": 0.0282, "step": 60680 }, { "epoch": 1.2127812637384596, "grad_norm": 0.11371560394763947, "learning_rate": 7.316877021335379e-05, "loss": 0.0319, "step": 60690 }, { "epoch": 1.2129810958794613, "grad_norm": 0.09771379083395004, "learning_rate": 7.315961195445932e-05, "loss": 0.0275, "step": 60700 }, { "epoch": 1.2131809280204628, "grad_norm": 0.14005526900291443, "learning_rate": 7.315045270622879e-05, "loss": 0.0331, "step": 60710 }, { "epoch": 1.2133807601614643, "grad_norm": 0.15346193313598633, "learning_rate": 7.314129246905346e-05, "loss": 0.0393, "step": 60720 }, { "epoch": 1.2135805923024658, "grad_norm": 0.15130825340747833, "learning_rate": 7.313213124332459e-05, "loss": 0.0386, "step": 60730 }, { "epoch": 1.2137804244434676, "grad_norm": 0.15106378495693207, "learning_rate": 7.31229690294336e-05, "loss": 0.0353, "step": 60740 }, { "epoch": 1.213980256584469, "grad_norm": 0.11679311841726303, "learning_rate": 7.311380582777184e-05, "loss": 0.0354, "step": 60750 }, { "epoch": 1.2141800887254706, "grad_norm": 0.1375085860490799, "learning_rate": 7.310464163873075e-05, "loss": 0.0359, "step": 60760 }, { "epoch": 1.214379920866472, "grad_norm": 0.14083190262317657, "learning_rate": 7.309547646270182e-05, "loss": 0.0363, "step": 60770 }, { "epoch": 1.2145797530074738, "grad_norm": 0.13128022849559784, "learning_rate": 7.308631030007657e-05, "loss": 0.0333, "step": 60780 }, { "epoch": 1.2147795851484753, "grad_norm": 0.15218888223171234, "learning_rate": 7.307714315124656e-05, "loss": 0.0325, "step": 60790 }, { "epoch": 1.2149794172894768, "grad_norm": 0.12562814354896545, "learning_rate": 7.306797501660335e-05, "loss": 0.0346, "step": 60800 }, { "epoch": 1.2151792494304785, "grad_norm": 0.13958460092544556, "learning_rate": 7.305880589653866e-05, "loss": 0.0306, "step": 60810 }, { "epoch": 1.21537908157148, "grad_norm": 0.1396370381116867, "learning_rate": 7.304963579144414e-05, "loss": 0.0324, "step": 60820 }, { "epoch": 1.2155789137124815, "grad_norm": 0.11332865059375763, "learning_rate": 7.304046470171151e-05, "loss": 0.0358, "step": 60830 }, { "epoch": 1.215778745853483, "grad_norm": 0.1266389936208725, "learning_rate": 7.303129262773256e-05, "loss": 0.0346, "step": 60840 }, { "epoch": 1.2159785779944847, "grad_norm": 0.12342233210802078, "learning_rate": 7.30221195698991e-05, "loss": 0.0369, "step": 60850 }, { "epoch": 1.2161784101354862, "grad_norm": 0.11478278040885925, "learning_rate": 7.301294552860299e-05, "loss": 0.0282, "step": 60860 }, { "epoch": 1.2163782422764877, "grad_norm": 0.1732473373413086, "learning_rate": 7.300377050423613e-05, "loss": 0.0389, "step": 60870 }, { "epoch": 1.2165780744174892, "grad_norm": 0.15521052479743958, "learning_rate": 7.299459449719042e-05, "loss": 0.0338, "step": 60880 }, { "epoch": 1.216777906558491, "grad_norm": 0.12036968022584915, "learning_rate": 7.29854175078579e-05, "loss": 0.0312, "step": 60890 }, { "epoch": 1.2169777386994924, "grad_norm": 0.13066203892230988, "learning_rate": 7.297623953663058e-05, "loss": 0.0338, "step": 60900 }, { "epoch": 1.217177570840494, "grad_norm": 0.2069413810968399, "learning_rate": 7.29670605839005e-05, "loss": 0.0337, "step": 60910 }, { "epoch": 1.2173774029814957, "grad_norm": 0.1305548995733261, "learning_rate": 7.295788065005978e-05, "loss": 0.0391, "step": 60920 }, { "epoch": 1.2175772351224972, "grad_norm": 0.12051037698984146, "learning_rate": 7.294869973550057e-05, "loss": 0.0336, "step": 60930 }, { "epoch": 1.2177770672634987, "grad_norm": 0.10796091705560684, "learning_rate": 7.293951784061509e-05, "loss": 0.0307, "step": 60940 }, { "epoch": 1.2179768994045002, "grad_norm": 0.12124072015285492, "learning_rate": 7.293033496579552e-05, "loss": 0.0291, "step": 60950 }, { "epoch": 1.2181767315455017, "grad_norm": 0.14496497809886932, "learning_rate": 7.292115111143419e-05, "loss": 0.0392, "step": 60960 }, { "epoch": 1.2183765636865034, "grad_norm": 0.12283980846405029, "learning_rate": 7.291196627792338e-05, "loss": 0.0348, "step": 60970 }, { "epoch": 1.2185763958275049, "grad_norm": 0.1329132616519928, "learning_rate": 7.290278046565546e-05, "loss": 0.0398, "step": 60980 }, { "epoch": 1.2187762279685064, "grad_norm": 0.12977087497711182, "learning_rate": 7.289359367502285e-05, "loss": 0.0297, "step": 60990 }, { "epoch": 1.218976060109508, "grad_norm": 0.13622091710567474, "learning_rate": 7.288440590641794e-05, "loss": 0.0333, "step": 61000 }, { "epoch": 1.2191758922505096, "grad_norm": 0.1896040290594101, "learning_rate": 7.287521716023327e-05, "loss": 0.0296, "step": 61010 }, { "epoch": 1.219375724391511, "grad_norm": 0.10867872834205627, "learning_rate": 7.286602743686135e-05, "loss": 0.0311, "step": 61020 }, { "epoch": 1.2195755565325126, "grad_norm": 0.13277243077754974, "learning_rate": 7.285683673669474e-05, "loss": 0.0369, "step": 61030 }, { "epoch": 1.2197753886735143, "grad_norm": 0.12346716225147247, "learning_rate": 7.284764506012604e-05, "loss": 0.0301, "step": 61040 }, { "epoch": 1.2199752208145158, "grad_norm": 0.13635991513729095, "learning_rate": 7.283845240754793e-05, "loss": 0.0397, "step": 61050 }, { "epoch": 1.2201750529555173, "grad_norm": 0.17125724256038666, "learning_rate": 7.282925877935309e-05, "loss": 0.037, "step": 61060 }, { "epoch": 1.2203748850965188, "grad_norm": 0.15445749461650848, "learning_rate": 7.282006417593426e-05, "loss": 0.0294, "step": 61070 }, { "epoch": 1.2205747172375205, "grad_norm": 0.12753728032112122, "learning_rate": 7.28108685976842e-05, "loss": 0.0365, "step": 61080 }, { "epoch": 1.220774549378522, "grad_norm": 0.1368575245141983, "learning_rate": 7.280167204499574e-05, "loss": 0.0304, "step": 61090 }, { "epoch": 1.2209743815195235, "grad_norm": 0.15130728483200073, "learning_rate": 7.279247451826174e-05, "loss": 0.042, "step": 61100 }, { "epoch": 1.2211742136605253, "grad_norm": 0.1049186959862709, "learning_rate": 7.278327601787508e-05, "loss": 0.0317, "step": 61110 }, { "epoch": 1.2213740458015268, "grad_norm": 0.19433066248893738, "learning_rate": 7.277407654422875e-05, "loss": 0.0376, "step": 61120 }, { "epoch": 1.2215738779425283, "grad_norm": 0.10764287412166595, "learning_rate": 7.276487609771569e-05, "loss": 0.0278, "step": 61130 }, { "epoch": 1.2217737100835298, "grad_norm": 0.15753217041492462, "learning_rate": 7.275567467872897e-05, "loss": 0.0396, "step": 61140 }, { "epoch": 1.2219735422245315, "grad_norm": 0.10046587884426117, "learning_rate": 7.27464722876616e-05, "loss": 0.0421, "step": 61150 }, { "epoch": 1.222173374365533, "grad_norm": 0.14244908094406128, "learning_rate": 7.273726892490674e-05, "loss": 0.0286, "step": 61160 }, { "epoch": 1.2223732065065345, "grad_norm": 0.13325883448123932, "learning_rate": 7.272806459085751e-05, "loss": 0.0363, "step": 61170 }, { "epoch": 1.222573038647536, "grad_norm": 0.1593448668718338, "learning_rate": 7.271885928590711e-05, "loss": 0.0247, "step": 61180 }, { "epoch": 1.2227728707885377, "grad_norm": 0.12408588826656342, "learning_rate": 7.27096530104488e-05, "loss": 0.0358, "step": 61190 }, { "epoch": 1.2229727029295392, "grad_norm": 0.1820707619190216, "learning_rate": 7.27004457648758e-05, "loss": 0.0403, "step": 61200 }, { "epoch": 1.2231725350705407, "grad_norm": 0.14126884937286377, "learning_rate": 7.269123754958149e-05, "loss": 0.0358, "step": 61210 }, { "epoch": 1.2233723672115424, "grad_norm": 0.1285441815853119, "learning_rate": 7.268202836495918e-05, "loss": 0.0281, "step": 61220 }, { "epoch": 1.223572199352544, "grad_norm": 0.12320858240127563, "learning_rate": 7.267281821140229e-05, "loss": 0.0342, "step": 61230 }, { "epoch": 1.2237720314935454, "grad_norm": 0.13428202271461487, "learning_rate": 7.266360708930425e-05, "loss": 0.024, "step": 61240 }, { "epoch": 1.223971863634547, "grad_norm": 0.12281098961830139, "learning_rate": 7.265439499905855e-05, "loss": 0.0281, "step": 61250 }, { "epoch": 1.2241716957755486, "grad_norm": 0.15437836945056915, "learning_rate": 7.264518194105874e-05, "loss": 0.0288, "step": 61260 }, { "epoch": 1.2243715279165501, "grad_norm": 0.14852185547351837, "learning_rate": 7.263596791569833e-05, "loss": 0.025, "step": 61270 }, { "epoch": 1.2245713600575516, "grad_norm": 0.10655441135168076, "learning_rate": 7.262675292337095e-05, "loss": 0.0275, "step": 61280 }, { "epoch": 1.2247711921985531, "grad_norm": 0.10106461495161057, "learning_rate": 7.261753696447027e-05, "loss": 0.0299, "step": 61290 }, { "epoch": 1.2249710243395548, "grad_norm": 0.09926902502775192, "learning_rate": 7.260832003938995e-05, "loss": 0.0277, "step": 61300 }, { "epoch": 1.2251708564805563, "grad_norm": 0.11086084693670273, "learning_rate": 7.259910214852373e-05, "loss": 0.0336, "step": 61310 }, { "epoch": 1.2253706886215578, "grad_norm": 0.09833910316228867, "learning_rate": 7.25898832922654e-05, "loss": 0.0333, "step": 61320 }, { "epoch": 1.2255705207625596, "grad_norm": 0.12375182658433914, "learning_rate": 7.258066347100874e-05, "loss": 0.0344, "step": 61330 }, { "epoch": 1.225770352903561, "grad_norm": 0.23885241150856018, "learning_rate": 7.257144268514762e-05, "loss": 0.0304, "step": 61340 }, { "epoch": 1.2259701850445626, "grad_norm": 0.12916965782642365, "learning_rate": 7.256222093507593e-05, "loss": 0.032, "step": 61350 }, { "epoch": 1.226170017185564, "grad_norm": 0.10518733412027359, "learning_rate": 7.255299822118761e-05, "loss": 0.0405, "step": 61360 }, { "epoch": 1.2263698493265656, "grad_norm": 0.11635612696409225, "learning_rate": 7.254377454387664e-05, "loss": 0.0272, "step": 61370 }, { "epoch": 1.2265696814675673, "grad_norm": 0.149875208735466, "learning_rate": 7.253454990353705e-05, "loss": 0.0346, "step": 61380 }, { "epoch": 1.2267695136085688, "grad_norm": 0.11963856965303421, "learning_rate": 7.252532430056287e-05, "loss": 0.0359, "step": 61390 }, { "epoch": 1.2269693457495703, "grad_norm": 0.1169508546590805, "learning_rate": 7.25160977353482e-05, "loss": 0.0277, "step": 61400 }, { "epoch": 1.227169177890572, "grad_norm": 0.13527940213680267, "learning_rate": 7.250687020828724e-05, "loss": 0.0395, "step": 61410 }, { "epoch": 1.2273690100315735, "grad_norm": 0.09971091896295547, "learning_rate": 7.249764171977409e-05, "loss": 0.0334, "step": 61420 }, { "epoch": 1.227568842172575, "grad_norm": 0.09598033130168915, "learning_rate": 7.248841227020304e-05, "loss": 0.0352, "step": 61430 }, { "epoch": 1.2277686743135765, "grad_norm": 0.14708398282527924, "learning_rate": 7.247918185996832e-05, "loss": 0.0319, "step": 61440 }, { "epoch": 1.2279685064545782, "grad_norm": 0.14133776724338531, "learning_rate": 7.246995048946425e-05, "loss": 0.0491, "step": 61450 }, { "epoch": 1.2281683385955797, "grad_norm": 0.14832361042499542, "learning_rate": 7.246071815908517e-05, "loss": 0.0332, "step": 61460 }, { "epoch": 1.2283681707365812, "grad_norm": 0.16395215690135956, "learning_rate": 7.245148486922549e-05, "loss": 0.0347, "step": 61470 }, { "epoch": 1.2285680028775827, "grad_norm": 0.12207159399986267, "learning_rate": 7.24422506202796e-05, "loss": 0.0321, "step": 61480 }, { "epoch": 1.2287678350185844, "grad_norm": 0.14172956347465515, "learning_rate": 7.243301541264199e-05, "loss": 0.0359, "step": 61490 }, { "epoch": 1.228967667159586, "grad_norm": 0.12872780859470367, "learning_rate": 7.242377924670718e-05, "loss": 0.0308, "step": 61500 }, { "epoch": 1.2291674993005874, "grad_norm": 0.14980053901672363, "learning_rate": 7.241454212286971e-05, "loss": 0.0305, "step": 61510 }, { "epoch": 1.2293673314415892, "grad_norm": 0.1171330064535141, "learning_rate": 7.240530404152418e-05, "loss": 0.0316, "step": 61520 }, { "epoch": 1.2295671635825907, "grad_norm": 0.12147774547338486, "learning_rate": 7.239606500306524e-05, "loss": 0.0312, "step": 61530 }, { "epoch": 1.2297669957235922, "grad_norm": 0.2062566578388214, "learning_rate": 7.238682500788752e-05, "loss": 0.044, "step": 61540 }, { "epoch": 1.2299668278645937, "grad_norm": 0.15000492334365845, "learning_rate": 7.237758405638578e-05, "loss": 0.0342, "step": 61550 }, { "epoch": 1.2301666600055954, "grad_norm": 0.11995617300271988, "learning_rate": 7.236834214895474e-05, "loss": 0.0327, "step": 61560 }, { "epoch": 1.2303664921465969, "grad_norm": 0.14171308279037476, "learning_rate": 7.235909928598922e-05, "loss": 0.0312, "step": 61570 }, { "epoch": 1.2305663242875984, "grad_norm": 0.1084805354475975, "learning_rate": 7.234985546788405e-05, "loss": 0.0408, "step": 61580 }, { "epoch": 1.2307661564285999, "grad_norm": 0.14562438428401947, "learning_rate": 7.234061069503412e-05, "loss": 0.0296, "step": 61590 }, { "epoch": 1.2309659885696016, "grad_norm": 0.1168879047036171, "learning_rate": 7.233136496783434e-05, "loss": 0.0314, "step": 61600 }, { "epoch": 1.231165820710603, "grad_norm": 0.13799606263637543, "learning_rate": 7.232211828667969e-05, "loss": 0.0335, "step": 61610 }, { "epoch": 1.2313656528516046, "grad_norm": 0.11269320547580719, "learning_rate": 7.231287065196512e-05, "loss": 0.0214, "step": 61620 }, { "epoch": 1.2315654849926063, "grad_norm": 0.12649036943912506, "learning_rate": 7.230362206408572e-05, "loss": 0.0318, "step": 61630 }, { "epoch": 1.2317653171336078, "grad_norm": 0.13686569035053253, "learning_rate": 7.229437252343659e-05, "loss": 0.0284, "step": 61640 }, { "epoch": 1.2319651492746093, "grad_norm": 0.13934235274791718, "learning_rate": 7.228512203041278e-05, "loss": 0.0387, "step": 61650 }, { "epoch": 1.2321649814156108, "grad_norm": 0.14338795840740204, "learning_rate": 7.227587058540952e-05, "loss": 0.0398, "step": 61660 }, { "epoch": 1.2323648135566125, "grad_norm": 0.14353229105472565, "learning_rate": 7.226661818882197e-05, "loss": 0.0353, "step": 61670 }, { "epoch": 1.232564645697614, "grad_norm": 0.11058422923088074, "learning_rate": 7.22573648410454e-05, "loss": 0.0324, "step": 61680 }, { "epoch": 1.2327644778386155, "grad_norm": 0.1374533474445343, "learning_rate": 7.22481105424751e-05, "loss": 0.0329, "step": 61690 }, { "epoch": 1.232964309979617, "grad_norm": 0.1795538067817688, "learning_rate": 7.22388552935064e-05, "loss": 0.0373, "step": 61700 }, { "epoch": 1.2331641421206188, "grad_norm": 0.13752321898937225, "learning_rate": 7.222959909453463e-05, "loss": 0.0321, "step": 61710 }, { "epoch": 1.2333639742616203, "grad_norm": 0.12605629861354828, "learning_rate": 7.222034194595524e-05, "loss": 0.0316, "step": 61720 }, { "epoch": 1.2335638064026218, "grad_norm": 0.1307329386472702, "learning_rate": 7.221108384816366e-05, "loss": 0.0307, "step": 61730 }, { "epoch": 1.2337636385436235, "grad_norm": 0.1479780524969101, "learning_rate": 7.220182480155538e-05, "loss": 0.0444, "step": 61740 }, { "epoch": 1.233963470684625, "grad_norm": 0.13014979660511017, "learning_rate": 7.219256480652593e-05, "loss": 0.0361, "step": 61750 }, { "epoch": 1.2341633028256265, "grad_norm": 0.12717628479003906, "learning_rate": 7.218330386347088e-05, "loss": 0.0345, "step": 61760 }, { "epoch": 1.234363134966628, "grad_norm": 0.15253432095050812, "learning_rate": 7.217404197278585e-05, "loss": 0.0305, "step": 61770 }, { "epoch": 1.2345629671076295, "grad_norm": 0.13258269429206848, "learning_rate": 7.216477913486647e-05, "loss": 0.0313, "step": 61780 }, { "epoch": 1.2347627992486312, "grad_norm": 0.09548202902078629, "learning_rate": 7.215551535010844e-05, "loss": 0.0312, "step": 61790 }, { "epoch": 1.2349626313896327, "grad_norm": 0.15290571749210358, "learning_rate": 7.214625061890751e-05, "loss": 0.0319, "step": 61800 }, { "epoch": 1.2351624635306342, "grad_norm": 0.12421216815710068, "learning_rate": 7.213698494165943e-05, "loss": 0.0322, "step": 61810 }, { "epoch": 1.235362295671636, "grad_norm": 0.13545389473438263, "learning_rate": 7.212771831876002e-05, "loss": 0.0376, "step": 61820 }, { "epoch": 1.2355621278126374, "grad_norm": 0.14488281309604645, "learning_rate": 7.211845075060514e-05, "loss": 0.0326, "step": 61830 }, { "epoch": 1.235761959953639, "grad_norm": 0.1596132516860962, "learning_rate": 7.210918223759068e-05, "loss": 0.0269, "step": 61840 }, { "epoch": 1.2359617920946404, "grad_norm": 0.1440214067697525, "learning_rate": 7.209991278011255e-05, "loss": 0.0288, "step": 61850 }, { "epoch": 1.2361616242356421, "grad_norm": 0.1317969560623169, "learning_rate": 7.209064237856676e-05, "loss": 0.0299, "step": 61860 }, { "epoch": 1.2363614563766436, "grad_norm": 0.1398826539516449, "learning_rate": 7.208137103334931e-05, "loss": 0.0277, "step": 61870 }, { "epoch": 1.2365612885176451, "grad_norm": 0.12584280967712402, "learning_rate": 7.207209874485624e-05, "loss": 0.0309, "step": 61880 }, { "epoch": 1.2367611206586466, "grad_norm": 0.12862060964107513, "learning_rate": 7.206282551348366e-05, "loss": 0.0419, "step": 61890 }, { "epoch": 1.2369609527996484, "grad_norm": 0.15137125551700592, "learning_rate": 7.205355133962772e-05, "loss": 0.0323, "step": 61900 }, { "epoch": 1.2371607849406498, "grad_norm": 0.157804474234581, "learning_rate": 7.20442762236846e-05, "loss": 0.0356, "step": 61910 }, { "epoch": 1.2373606170816513, "grad_norm": 0.1562696099281311, "learning_rate": 7.203500016605045e-05, "loss": 0.032, "step": 61920 }, { "epoch": 1.237560449222653, "grad_norm": 0.11924010515213013, "learning_rate": 7.202572316712159e-05, "loss": 0.0376, "step": 61930 }, { "epoch": 1.2377602813636546, "grad_norm": 0.14207880198955536, "learning_rate": 7.201644522729432e-05, "loss": 0.0365, "step": 61940 }, { "epoch": 1.237960113504656, "grad_norm": 0.10886427015066147, "learning_rate": 7.200716634696493e-05, "loss": 0.0367, "step": 61950 }, { "epoch": 1.2381599456456576, "grad_norm": 0.11932792514562607, "learning_rate": 7.199788652652982e-05, "loss": 0.032, "step": 61960 }, { "epoch": 1.2383597777866593, "grad_norm": 0.15708845853805542, "learning_rate": 7.198860576638542e-05, "loss": 0.034, "step": 61970 }, { "epoch": 1.2385596099276608, "grad_norm": 0.17771880328655243, "learning_rate": 7.197932406692818e-05, "loss": 0.0353, "step": 61980 }, { "epoch": 1.2387594420686623, "grad_norm": 0.14102187752723694, "learning_rate": 7.197004142855459e-05, "loss": 0.0386, "step": 61990 }, { "epoch": 1.2389592742096638, "grad_norm": 0.12521594762802124, "learning_rate": 7.196075785166118e-05, "loss": 0.0349, "step": 62000 }, { "epoch": 1.2391591063506655, "grad_norm": 0.1220841258764267, "learning_rate": 7.195147333664455e-05, "loss": 0.0336, "step": 62010 }, { "epoch": 1.239358938491667, "grad_norm": 0.12079732865095139, "learning_rate": 7.19421878839013e-05, "loss": 0.0294, "step": 62020 }, { "epoch": 1.2395587706326685, "grad_norm": 0.1407458484172821, "learning_rate": 7.19329014938281e-05, "loss": 0.031, "step": 62030 }, { "epoch": 1.2397586027736702, "grad_norm": 0.12993913888931274, "learning_rate": 7.192361416682164e-05, "loss": 0.0368, "step": 62040 }, { "epoch": 1.2399584349146717, "grad_norm": 0.14879557490348816, "learning_rate": 7.191432590327864e-05, "loss": 0.0373, "step": 62050 }, { "epoch": 1.2401582670556732, "grad_norm": 0.16529078781604767, "learning_rate": 7.190503670359592e-05, "loss": 0.0404, "step": 62060 }, { "epoch": 1.2403580991966747, "grad_norm": 0.15617074072360992, "learning_rate": 7.189574656817026e-05, "loss": 0.0312, "step": 62070 }, { "epoch": 1.2405579313376764, "grad_norm": 0.12436247617006302, "learning_rate": 7.188645549739852e-05, "loss": 0.0373, "step": 62080 }, { "epoch": 1.240757763478678, "grad_norm": 0.16023071110248566, "learning_rate": 7.187716349167764e-05, "loss": 0.0278, "step": 62090 }, { "epoch": 1.2409575956196794, "grad_norm": 0.12437178194522858, "learning_rate": 7.186787055140448e-05, "loss": 0.0299, "step": 62100 }, { "epoch": 1.241157427760681, "grad_norm": 0.11796342581510544, "learning_rate": 7.18585766769761e-05, "loss": 0.0284, "step": 62110 }, { "epoch": 1.2413572599016827, "grad_norm": 0.12321551144123077, "learning_rate": 7.184928186878945e-05, "loss": 0.0371, "step": 62120 }, { "epoch": 1.2415570920426842, "grad_norm": 0.1480621099472046, "learning_rate": 7.183998612724165e-05, "loss": 0.0354, "step": 62130 }, { "epoch": 1.2417569241836857, "grad_norm": 0.14284738898277283, "learning_rate": 7.183068945272974e-05, "loss": 0.0363, "step": 62140 }, { "epoch": 1.2419567563246874, "grad_norm": 0.15654684603214264, "learning_rate": 7.182139184565089e-05, "loss": 0.0352, "step": 62150 }, { "epoch": 1.2421565884656889, "grad_norm": 0.11924179643392563, "learning_rate": 7.181209330640226e-05, "loss": 0.0314, "step": 62160 }, { "epoch": 1.2423564206066904, "grad_norm": 0.11417710036039352, "learning_rate": 7.180279383538107e-05, "loss": 0.0369, "step": 62170 }, { "epoch": 1.2425562527476919, "grad_norm": 0.12219561636447906, "learning_rate": 7.17934934329846e-05, "loss": 0.0425, "step": 62180 }, { "epoch": 1.2427560848886934, "grad_norm": 0.13223116099834442, "learning_rate": 7.178419209961012e-05, "loss": 0.0355, "step": 62190 }, { "epoch": 1.242955917029695, "grad_norm": 0.08292356878519058, "learning_rate": 7.177488983565497e-05, "loss": 0.0268, "step": 62200 }, { "epoch": 1.2431557491706966, "grad_norm": 0.12479452043771744, "learning_rate": 7.176558664151652e-05, "loss": 0.042, "step": 62210 }, { "epoch": 1.243355581311698, "grad_norm": 0.14484462141990662, "learning_rate": 7.17562825175922e-05, "loss": 0.0346, "step": 62220 }, { "epoch": 1.2435554134526998, "grad_norm": 0.09504485875368118, "learning_rate": 7.174697746427945e-05, "loss": 0.0378, "step": 62230 }, { "epoch": 1.2437552455937013, "grad_norm": 0.10201051831245422, "learning_rate": 7.17376714819758e-05, "loss": 0.0272, "step": 62240 }, { "epoch": 1.2439550777347028, "grad_norm": 0.14128954708576202, "learning_rate": 7.172836457107872e-05, "loss": 0.0348, "step": 62250 }, { "epoch": 1.2441549098757043, "grad_norm": 0.14372487366199493, "learning_rate": 7.171905673198586e-05, "loss": 0.0386, "step": 62260 }, { "epoch": 1.244354742016706, "grad_norm": 0.15861259400844574, "learning_rate": 7.170974796509478e-05, "loss": 0.0412, "step": 62270 }, { "epoch": 1.2445545741577075, "grad_norm": 0.11468281596899033, "learning_rate": 7.170043827080313e-05, "loss": 0.0341, "step": 62280 }, { "epoch": 1.244754406298709, "grad_norm": 0.1412515789270401, "learning_rate": 7.169112764950865e-05, "loss": 0.0381, "step": 62290 }, { "epoch": 1.2449542384397105, "grad_norm": 0.19266019761562347, "learning_rate": 7.168181610160903e-05, "loss": 0.0384, "step": 62300 }, { "epoch": 1.2451540705807123, "grad_norm": 0.12078236788511276, "learning_rate": 7.167250362750206e-05, "loss": 0.0327, "step": 62310 }, { "epoch": 1.2453539027217138, "grad_norm": 0.12998870015144348, "learning_rate": 7.166319022758554e-05, "loss": 0.031, "step": 62320 }, { "epoch": 1.2455537348627153, "grad_norm": 0.1546170860528946, "learning_rate": 7.165387590225734e-05, "loss": 0.0245, "step": 62330 }, { "epoch": 1.245753567003717, "grad_norm": 0.15889135003089905, "learning_rate": 7.164456065191534e-05, "loss": 0.0347, "step": 62340 }, { "epoch": 1.2459533991447185, "grad_norm": 0.16513574123382568, "learning_rate": 7.163524447695746e-05, "loss": 0.0316, "step": 62350 }, { "epoch": 1.24615323128572, "grad_norm": 0.15282845497131348, "learning_rate": 7.16259273777817e-05, "loss": 0.0282, "step": 62360 }, { "epoch": 1.2463530634267215, "grad_norm": 0.13374583423137665, "learning_rate": 7.161660935478604e-05, "loss": 0.0356, "step": 62370 }, { "epoch": 1.2465528955677232, "grad_norm": 0.17265239357948303, "learning_rate": 7.160729040836856e-05, "loss": 0.0303, "step": 62380 }, { "epoch": 1.2467527277087247, "grad_norm": 0.17651979625225067, "learning_rate": 7.15979705389273e-05, "loss": 0.0354, "step": 62390 }, { "epoch": 1.2469525598497262, "grad_norm": 0.1770825833082199, "learning_rate": 7.158864974686042e-05, "loss": 0.0313, "step": 62400 }, { "epoch": 1.2471523919907277, "grad_norm": 0.14241255819797516, "learning_rate": 7.157932803256609e-05, "loss": 0.0293, "step": 62410 }, { "epoch": 1.2473522241317294, "grad_norm": 0.13809412717819214, "learning_rate": 7.15700053964425e-05, "loss": 0.0296, "step": 62420 }, { "epoch": 1.247552056272731, "grad_norm": 0.11368552595376968, "learning_rate": 7.156068183888791e-05, "loss": 0.0347, "step": 62430 }, { "epoch": 1.2477518884137324, "grad_norm": 0.12390701472759247, "learning_rate": 7.15513573603006e-05, "loss": 0.0427, "step": 62440 }, { "epoch": 1.2479517205547341, "grad_norm": 0.1307661533355713, "learning_rate": 7.15420319610789e-05, "loss": 0.0267, "step": 62450 }, { "epoch": 1.2481515526957356, "grad_norm": 0.11784256994724274, "learning_rate": 7.153270564162116e-05, "loss": 0.0277, "step": 62460 }, { "epoch": 1.2483513848367371, "grad_norm": 0.14627081155776978, "learning_rate": 7.15233784023258e-05, "loss": 0.0374, "step": 62470 }, { "epoch": 1.2485512169777386, "grad_norm": 0.11401718109846115, "learning_rate": 7.151405024359125e-05, "loss": 0.0302, "step": 62480 }, { "epoch": 1.2487510491187404, "grad_norm": 0.14650073647499084, "learning_rate": 7.1504721165816e-05, "loss": 0.0294, "step": 62490 }, { "epoch": 1.2489508812597419, "grad_norm": 0.12819062173366547, "learning_rate": 7.149539116939855e-05, "loss": 0.0337, "step": 62500 }, { "epoch": 1.2491507134007434, "grad_norm": 0.1191619336605072, "learning_rate": 7.148606025473749e-05, "loss": 0.0285, "step": 62510 }, { "epoch": 1.2493505455417449, "grad_norm": 0.12672090530395508, "learning_rate": 7.14767284222314e-05, "loss": 0.0291, "step": 62520 }, { "epoch": 1.2495503776827466, "grad_norm": 0.12222599238157272, "learning_rate": 7.146739567227896e-05, "loss": 0.0331, "step": 62530 }, { "epoch": 1.249750209823748, "grad_norm": 0.11089076101779938, "learning_rate": 7.145806200527878e-05, "loss": 0.0319, "step": 62540 }, { "epoch": 1.2499500419647496, "grad_norm": 0.08819796144962311, "learning_rate": 7.144872742162962e-05, "loss": 0.0303, "step": 62550 }, { "epoch": 1.2501498741057513, "grad_norm": 0.19689248502254486, "learning_rate": 7.143939192173022e-05, "loss": 0.0345, "step": 62560 }, { "epoch": 1.2503497062467528, "grad_norm": 0.13968758285045624, "learning_rate": 7.143005550597938e-05, "loss": 0.0339, "step": 62570 }, { "epoch": 1.2505495383877543, "grad_norm": 0.13349123299121857, "learning_rate": 7.142071817477596e-05, "loss": 0.0351, "step": 62580 }, { "epoch": 1.2507493705287558, "grad_norm": 0.12012631446123123, "learning_rate": 7.141137992851879e-05, "loss": 0.0297, "step": 62590 }, { "epoch": 1.2509492026697573, "grad_norm": 0.1109909638762474, "learning_rate": 7.140204076760681e-05, "loss": 0.0296, "step": 62600 }, { "epoch": 1.251149034810759, "grad_norm": 0.13142229616641998, "learning_rate": 7.139270069243896e-05, "loss": 0.0324, "step": 62610 }, { "epoch": 1.2513488669517605, "grad_norm": 0.12696176767349243, "learning_rate": 7.138335970341425e-05, "loss": 0.025, "step": 62620 }, { "epoch": 1.251548699092762, "grad_norm": 0.17355674505233765, "learning_rate": 7.137401780093168e-05, "loss": 0.0318, "step": 62630 }, { "epoch": 1.2517485312337637, "grad_norm": 0.13306686282157898, "learning_rate": 7.136467498539035e-05, "loss": 0.0458, "step": 62640 }, { "epoch": 1.2519483633747652, "grad_norm": 0.16196738183498383, "learning_rate": 7.135533125718935e-05, "loss": 0.0309, "step": 62650 }, { "epoch": 1.2521481955157667, "grad_norm": 0.15636852383613586, "learning_rate": 7.134598661672781e-05, "loss": 0.0317, "step": 62660 }, { "epoch": 1.2523480276567684, "grad_norm": 0.16150617599487305, "learning_rate": 7.133664106440495e-05, "loss": 0.0307, "step": 62670 }, { "epoch": 1.25254785979777, "grad_norm": 0.16524310410022736, "learning_rate": 7.132729460061997e-05, "loss": 0.03, "step": 62680 }, { "epoch": 1.2527476919387714, "grad_norm": 0.1242363303899765, "learning_rate": 7.131794722577218e-05, "loss": 0.0248, "step": 62690 }, { "epoch": 1.252947524079773, "grad_norm": 0.12978966534137726, "learning_rate": 7.13085989402608e-05, "loss": 0.0308, "step": 62700 }, { "epoch": 1.2531473562207744, "grad_norm": 0.16770508885383606, "learning_rate": 7.129924974448525e-05, "loss": 0.0343, "step": 62710 }, { "epoch": 1.2533471883617762, "grad_norm": 0.13952703773975372, "learning_rate": 7.128989963884485e-05, "loss": 0.0321, "step": 62720 }, { "epoch": 1.2535470205027777, "grad_norm": 0.1464841067790985, "learning_rate": 7.128054862373907e-05, "loss": 0.0291, "step": 62730 }, { "epoch": 1.2537468526437792, "grad_norm": 0.1570199429988861, "learning_rate": 7.127119669956735e-05, "loss": 0.031, "step": 62740 }, { "epoch": 1.2539466847847809, "grad_norm": 0.12409797310829163, "learning_rate": 7.126184386672919e-05, "loss": 0.0331, "step": 62750 }, { "epoch": 1.2541465169257824, "grad_norm": 0.15936487913131714, "learning_rate": 7.125249012562409e-05, "loss": 0.0275, "step": 62760 }, { "epoch": 1.2543463490667839, "grad_norm": 0.1279560774564743, "learning_rate": 7.124313547665166e-05, "loss": 0.0332, "step": 62770 }, { "epoch": 1.2545461812077854, "grad_norm": 0.15888750553131104, "learning_rate": 7.123377992021153e-05, "loss": 0.032, "step": 62780 }, { "epoch": 1.2547460133487869, "grad_norm": 0.1377919465303421, "learning_rate": 7.122442345670332e-05, "loss": 0.0314, "step": 62790 }, { "epoch": 1.2549458454897886, "grad_norm": 0.0996474027633667, "learning_rate": 7.121506608652673e-05, "loss": 0.0347, "step": 62800 }, { "epoch": 1.25514567763079, "grad_norm": 0.11531158536672592, "learning_rate": 7.120570781008148e-05, "loss": 0.0391, "step": 62810 }, { "epoch": 1.2553455097717916, "grad_norm": 0.140184223651886, "learning_rate": 7.119634862776735e-05, "loss": 0.0318, "step": 62820 }, { "epoch": 1.2555453419127933, "grad_norm": 0.17166000604629517, "learning_rate": 7.118698853998416e-05, "loss": 0.0255, "step": 62830 }, { "epoch": 1.2557451740537948, "grad_norm": 0.1369025558233261, "learning_rate": 7.117762754713173e-05, "loss": 0.0346, "step": 62840 }, { "epoch": 1.2559450061947963, "grad_norm": 0.16660374402999878, "learning_rate": 7.116826564960996e-05, "loss": 0.0327, "step": 62850 }, { "epoch": 1.256144838335798, "grad_norm": 0.11094950884580612, "learning_rate": 7.115890284781876e-05, "loss": 0.0238, "step": 62860 }, { "epoch": 1.2563446704767995, "grad_norm": 0.14763931930065155, "learning_rate": 7.11495391421581e-05, "loss": 0.0377, "step": 62870 }, { "epoch": 1.256544502617801, "grad_norm": 0.11837849766016006, "learning_rate": 7.114017453302799e-05, "loss": 0.0288, "step": 62880 }, { "epoch": 1.2567443347588025, "grad_norm": 0.15021762251853943, "learning_rate": 7.113080902082847e-05, "loss": 0.0344, "step": 62890 }, { "epoch": 1.256944166899804, "grad_norm": 0.13480688631534576, "learning_rate": 7.112144260595959e-05, "loss": 0.0261, "step": 62900 }, { "epoch": 1.2571439990408058, "grad_norm": 0.11687096953392029, "learning_rate": 7.111207528882149e-05, "loss": 0.0312, "step": 62910 }, { "epoch": 1.2573438311818073, "grad_norm": 0.12246131896972656, "learning_rate": 7.11027070698143e-05, "loss": 0.0317, "step": 62920 }, { "epoch": 1.2575436633228088, "grad_norm": 0.14359793066978455, "learning_rate": 7.109333794933825e-05, "loss": 0.0287, "step": 62930 }, { "epoch": 1.2577434954638105, "grad_norm": 0.1437821239233017, "learning_rate": 7.108396792779356e-05, "loss": 0.038, "step": 62940 }, { "epoch": 1.257943327604812, "grad_norm": 0.11679224669933319, "learning_rate": 7.107459700558047e-05, "loss": 0.0253, "step": 62950 }, { "epoch": 1.2581431597458135, "grad_norm": 0.1410655528306961, "learning_rate": 7.106522518309934e-05, "loss": 0.0311, "step": 62960 }, { "epoch": 1.2583429918868152, "grad_norm": 0.11248014122247696, "learning_rate": 7.105585246075046e-05, "loss": 0.0321, "step": 62970 }, { "epoch": 1.2585428240278167, "grad_norm": 0.1519182175397873, "learning_rate": 7.104647883893427e-05, "loss": 0.0403, "step": 62980 }, { "epoch": 1.2587426561688182, "grad_norm": 0.10427774488925934, "learning_rate": 7.103710431805115e-05, "loss": 0.0367, "step": 62990 }, { "epoch": 1.2589424883098197, "grad_norm": 0.1455468386411667, "learning_rate": 7.102772889850158e-05, "loss": 0.0362, "step": 63000 }, { "epoch": 1.2591423204508212, "grad_norm": 0.11849676817655563, "learning_rate": 7.101835258068606e-05, "loss": 0.0389, "step": 63010 }, { "epoch": 1.259342152591823, "grad_norm": 0.1341399997472763, "learning_rate": 7.100897536500512e-05, "loss": 0.0316, "step": 63020 }, { "epoch": 1.2595419847328244, "grad_norm": 0.13133184611797333, "learning_rate": 7.099959725185937e-05, "loss": 0.0269, "step": 63030 }, { "epoch": 1.259741816873826, "grad_norm": 0.1209232434630394, "learning_rate": 7.09902182416494e-05, "loss": 0.0283, "step": 63040 }, { "epoch": 1.2599416490148276, "grad_norm": 0.12629827857017517, "learning_rate": 7.098083833477585e-05, "loss": 0.0323, "step": 63050 }, { "epoch": 1.2601414811558291, "grad_norm": 0.1749918907880783, "learning_rate": 7.097145753163944e-05, "loss": 0.0376, "step": 63060 }, { "epoch": 1.2603413132968306, "grad_norm": 0.15254923701286316, "learning_rate": 7.09620758326409e-05, "loss": 0.037, "step": 63070 }, { "epoch": 1.2605411454378324, "grad_norm": 0.10226918011903763, "learning_rate": 7.095269323818097e-05, "loss": 0.0294, "step": 63080 }, { "epoch": 1.2607409775788339, "grad_norm": 0.1462079882621765, "learning_rate": 7.094330974866049e-05, "loss": 0.0271, "step": 63090 }, { "epoch": 1.2609408097198354, "grad_norm": 0.23895147442817688, "learning_rate": 7.093392536448027e-05, "loss": 0.0325, "step": 63100 }, { "epoch": 1.2611406418608369, "grad_norm": 0.1353832483291626, "learning_rate": 7.092454008604125e-05, "loss": 0.0443, "step": 63110 }, { "epoch": 1.2613404740018384, "grad_norm": 0.15138396620750427, "learning_rate": 7.091515391374429e-05, "loss": 0.033, "step": 63120 }, { "epoch": 1.26154030614284, "grad_norm": 0.11342383176088333, "learning_rate": 7.090576684799038e-05, "loss": 0.0377, "step": 63130 }, { "epoch": 1.2617401382838416, "grad_norm": 0.11665698885917664, "learning_rate": 7.089637888918052e-05, "loss": 0.0435, "step": 63140 }, { "epoch": 1.261939970424843, "grad_norm": 0.14643451571464539, "learning_rate": 7.088699003771575e-05, "loss": 0.0412, "step": 63150 }, { "epoch": 1.2621398025658448, "grad_norm": 0.12130513787269592, "learning_rate": 7.087760029399713e-05, "loss": 0.0302, "step": 63160 }, { "epoch": 1.2623396347068463, "grad_norm": 0.1320977658033371, "learning_rate": 7.086820965842577e-05, "loss": 0.0267, "step": 63170 }, { "epoch": 1.2625394668478478, "grad_norm": 0.09341400116682053, "learning_rate": 7.085881813140285e-05, "loss": 0.0341, "step": 63180 }, { "epoch": 1.2627392989888493, "grad_norm": 0.1450244039297104, "learning_rate": 7.084942571332952e-05, "loss": 0.0377, "step": 63190 }, { "epoch": 1.2629391311298508, "grad_norm": 0.16561178863048553, "learning_rate": 7.084003240460703e-05, "loss": 0.0472, "step": 63200 }, { "epoch": 1.2631389632708525, "grad_norm": 0.10081014037132263, "learning_rate": 7.083063820563665e-05, "loss": 0.0307, "step": 63210 }, { "epoch": 1.263338795411854, "grad_norm": 0.12771423161029816, "learning_rate": 7.082124311681965e-05, "loss": 0.044, "step": 63220 }, { "epoch": 1.2635386275528555, "grad_norm": 0.13090018928050995, "learning_rate": 7.081184713855742e-05, "loss": 0.0257, "step": 63230 }, { "epoch": 1.2637384596938572, "grad_norm": 0.08778441697359085, "learning_rate": 7.080245027125129e-05, "loss": 0.0339, "step": 63240 }, { "epoch": 1.2639382918348587, "grad_norm": 0.1381208747625351, "learning_rate": 7.079305251530272e-05, "loss": 0.0346, "step": 63250 }, { "epoch": 1.2641381239758602, "grad_norm": 0.13073569536209106, "learning_rate": 7.078365387111311e-05, "loss": 0.0291, "step": 63260 }, { "epoch": 1.264337956116862, "grad_norm": 0.14389246702194214, "learning_rate": 7.077425433908402e-05, "loss": 0.0359, "step": 63270 }, { "epoch": 1.2645377882578634, "grad_norm": 0.11849981546401978, "learning_rate": 7.076485391961694e-05, "loss": 0.0333, "step": 63280 }, { "epoch": 1.264737620398865, "grad_norm": 0.12776342034339905, "learning_rate": 7.075545261311346e-05, "loss": 0.0255, "step": 63290 }, { "epoch": 1.2649374525398664, "grad_norm": 0.12480399012565613, "learning_rate": 7.074605041997516e-05, "loss": 0.0327, "step": 63300 }, { "epoch": 1.265137284680868, "grad_norm": 0.14567871391773224, "learning_rate": 7.07366473406037e-05, "loss": 0.0317, "step": 63310 }, { "epoch": 1.2653371168218697, "grad_norm": 0.12367493659257889, "learning_rate": 7.072724337540075e-05, "loss": 0.0347, "step": 63320 }, { "epoch": 1.2655369489628712, "grad_norm": 0.11649171262979507, "learning_rate": 7.071783852476805e-05, "loss": 0.0377, "step": 63330 }, { "epoch": 1.2657367811038727, "grad_norm": 0.1283174753189087, "learning_rate": 7.070843278910734e-05, "loss": 0.0371, "step": 63340 }, { "epoch": 1.2659366132448744, "grad_norm": 0.14804138243198395, "learning_rate": 7.069902616882044e-05, "loss": 0.0279, "step": 63350 }, { "epoch": 1.2661364453858759, "grad_norm": 0.1302609145641327, "learning_rate": 7.068961866430917e-05, "loss": 0.0269, "step": 63360 }, { "epoch": 1.2663362775268774, "grad_norm": 0.13793081045150757, "learning_rate": 7.068021027597536e-05, "loss": 0.0444, "step": 63370 }, { "epoch": 1.266536109667879, "grad_norm": 0.11202213913202286, "learning_rate": 7.0670801004221e-05, "loss": 0.0362, "step": 63380 }, { "epoch": 1.2667359418088806, "grad_norm": 0.12857495248317719, "learning_rate": 7.066139084944799e-05, "loss": 0.0309, "step": 63390 }, { "epoch": 1.266935773949882, "grad_norm": 0.12399362027645111, "learning_rate": 7.065197981205828e-05, "loss": 0.0317, "step": 63400 }, { "epoch": 1.2671356060908836, "grad_norm": 0.14930307865142822, "learning_rate": 7.064256789245399e-05, "loss": 0.0273, "step": 63410 }, { "epoch": 1.267335438231885, "grad_norm": 0.13123425841331482, "learning_rate": 7.063315509103707e-05, "loss": 0.045, "step": 63420 }, { "epoch": 1.2675352703728868, "grad_norm": 0.13776646554470062, "learning_rate": 7.062374140820969e-05, "loss": 0.0436, "step": 63430 }, { "epoch": 1.2677351025138883, "grad_norm": 0.14429646730422974, "learning_rate": 7.061432684437397e-05, "loss": 0.0335, "step": 63440 }, { "epoch": 1.2679349346548898, "grad_norm": 0.18895019590854645, "learning_rate": 7.060491139993208e-05, "loss": 0.0322, "step": 63450 }, { "epoch": 1.2681347667958915, "grad_norm": 0.13349202275276184, "learning_rate": 7.05954950752862e-05, "loss": 0.0348, "step": 63460 }, { "epoch": 1.268334598936893, "grad_norm": 0.1598004549741745, "learning_rate": 7.058607787083863e-05, "loss": 0.0308, "step": 63470 }, { "epoch": 1.2685344310778945, "grad_norm": 0.12851284444332123, "learning_rate": 7.057665978699161e-05, "loss": 0.0348, "step": 63480 }, { "epoch": 1.2687342632188963, "grad_norm": 0.09779471904039383, "learning_rate": 7.056724082414752e-05, "loss": 0.0319, "step": 63490 }, { "epoch": 1.2689340953598978, "grad_norm": 0.10297826677560806, "learning_rate": 7.055782098270866e-05, "loss": 0.0328, "step": 63500 }, { "epoch": 1.2691339275008993, "grad_norm": 0.11184831708669662, "learning_rate": 7.054840026307745e-05, "loss": 0.0319, "step": 63510 }, { "epoch": 1.2693337596419008, "grad_norm": 0.11369675397872925, "learning_rate": 7.053897866565634e-05, "loss": 0.0333, "step": 63520 }, { "epoch": 1.2695335917829023, "grad_norm": 0.1320866048336029, "learning_rate": 7.052955619084779e-05, "loss": 0.0308, "step": 63530 }, { "epoch": 1.269733423923904, "grad_norm": 0.13319386541843414, "learning_rate": 7.052013283905431e-05, "loss": 0.0349, "step": 63540 }, { "epoch": 1.2699332560649055, "grad_norm": 0.14103178679943085, "learning_rate": 7.051070861067845e-05, "loss": 0.0345, "step": 63550 }, { "epoch": 1.270133088205907, "grad_norm": 0.17165623605251312, "learning_rate": 7.050128350612282e-05, "loss": 0.049, "step": 63560 }, { "epoch": 1.2703329203469087, "grad_norm": 0.17268745601177216, "learning_rate": 7.049185752579e-05, "loss": 0.0308, "step": 63570 }, { "epoch": 1.2705327524879102, "grad_norm": 0.1281554102897644, "learning_rate": 7.048243067008267e-05, "loss": 0.0299, "step": 63580 }, { "epoch": 1.2707325846289117, "grad_norm": 0.09990691393613815, "learning_rate": 7.047300293940354e-05, "loss": 0.0366, "step": 63590 }, { "epoch": 1.2709324167699132, "grad_norm": 0.1361694186925888, "learning_rate": 7.046357433415533e-05, "loss": 0.0305, "step": 63600 }, { "epoch": 1.2711322489109147, "grad_norm": 0.13704344630241394, "learning_rate": 7.04541448547408e-05, "loss": 0.0272, "step": 63610 }, { "epoch": 1.2713320810519164, "grad_norm": 0.11019247770309448, "learning_rate": 7.04447145015628e-05, "loss": 0.0265, "step": 63620 }, { "epoch": 1.271531913192918, "grad_norm": 0.1155412569642067, "learning_rate": 7.043528327502415e-05, "loss": 0.0372, "step": 63630 }, { "epoch": 1.2717317453339194, "grad_norm": 0.14190609753131866, "learning_rate": 7.042585117552773e-05, "loss": 0.0416, "step": 63640 }, { "epoch": 1.2719315774749211, "grad_norm": 0.12491559237241745, "learning_rate": 7.041641820347649e-05, "loss": 0.0319, "step": 63650 }, { "epoch": 1.2721314096159226, "grad_norm": 0.14893434941768646, "learning_rate": 7.040698435927335e-05, "loss": 0.0312, "step": 63660 }, { "epoch": 1.2723312417569241, "grad_norm": 0.19578011333942413, "learning_rate": 7.039754964332133e-05, "loss": 0.0334, "step": 63670 }, { "epoch": 1.2725310738979259, "grad_norm": 0.14436030387878418, "learning_rate": 7.038811405602347e-05, "loss": 0.0372, "step": 63680 }, { "epoch": 1.2727309060389274, "grad_norm": 0.13139159977436066, "learning_rate": 7.037867759778284e-05, "loss": 0.0351, "step": 63690 }, { "epoch": 1.2729307381799289, "grad_norm": 0.13568831980228424, "learning_rate": 7.036924026900252e-05, "loss": 0.0276, "step": 63700 }, { "epoch": 1.2731305703209304, "grad_norm": 0.1878674179315567, "learning_rate": 7.035980207008568e-05, "loss": 0.0263, "step": 63710 }, { "epoch": 1.2733304024619319, "grad_norm": 0.14485584199428558, "learning_rate": 7.03503630014355e-05, "loss": 0.0359, "step": 63720 }, { "epoch": 1.2735302346029336, "grad_norm": 0.13473626971244812, "learning_rate": 7.03409230634552e-05, "loss": 0.0352, "step": 63730 }, { "epoch": 1.273730066743935, "grad_norm": 0.12095190584659576, "learning_rate": 7.033148225654804e-05, "loss": 0.0296, "step": 63740 }, { "epoch": 1.2739298988849366, "grad_norm": 0.11837572604417801, "learning_rate": 7.032204058111729e-05, "loss": 0.0243, "step": 63750 }, { "epoch": 1.2741297310259383, "grad_norm": 0.11523860692977905, "learning_rate": 7.031259803756631e-05, "loss": 0.0354, "step": 63760 }, { "epoch": 1.2743295631669398, "grad_norm": 0.14596262574195862, "learning_rate": 7.030315462629847e-05, "loss": 0.0291, "step": 63770 }, { "epoch": 1.2745293953079413, "grad_norm": 0.14237122237682343, "learning_rate": 7.029371034771714e-05, "loss": 0.0343, "step": 63780 }, { "epoch": 1.274729227448943, "grad_norm": 0.1065315306186676, "learning_rate": 7.02842652022258e-05, "loss": 0.0406, "step": 63790 }, { "epoch": 1.2749290595899445, "grad_norm": 0.13935957849025726, "learning_rate": 7.02748191902279e-05, "loss": 0.04, "step": 63800 }, { "epoch": 1.275128891730946, "grad_norm": 0.1471686065196991, "learning_rate": 7.026537231212699e-05, "loss": 0.0268, "step": 63810 }, { "epoch": 1.2753287238719475, "grad_norm": 0.13416913151741028, "learning_rate": 7.025592456832657e-05, "loss": 0.0352, "step": 63820 }, { "epoch": 1.275528556012949, "grad_norm": 0.17606161534786224, "learning_rate": 7.024647595923031e-05, "loss": 0.0304, "step": 63830 }, { "epoch": 1.2757283881539507, "grad_norm": 0.14226841926574707, "learning_rate": 7.023702648524175e-05, "loss": 0.0323, "step": 63840 }, { "epoch": 1.2759282202949522, "grad_norm": 0.11314777284860611, "learning_rate": 7.022757614676462e-05, "loss": 0.0284, "step": 63850 }, { "epoch": 1.2761280524359537, "grad_norm": 0.12702371180057526, "learning_rate": 7.021812494420259e-05, "loss": 0.0338, "step": 63860 }, { "epoch": 1.2763278845769555, "grad_norm": 0.13012175261974335, "learning_rate": 7.02086728779594e-05, "loss": 0.0409, "step": 63870 }, { "epoch": 1.276527716717957, "grad_norm": 0.1659204512834549, "learning_rate": 7.019921994843883e-05, "loss": 0.0375, "step": 63880 }, { "epoch": 1.2767275488589584, "grad_norm": 0.10632314532995224, "learning_rate": 7.01897661560447e-05, "loss": 0.0294, "step": 63890 }, { "epoch": 1.2769273809999602, "grad_norm": 0.1003803163766861, "learning_rate": 7.018031150118083e-05, "loss": 0.0285, "step": 63900 }, { "epoch": 1.2771272131409617, "grad_norm": 0.14935170114040375, "learning_rate": 7.017085598425113e-05, "loss": 0.041, "step": 63910 }, { "epoch": 1.2773270452819632, "grad_norm": 0.1433633714914322, "learning_rate": 7.016139960565952e-05, "loss": 0.0414, "step": 63920 }, { "epoch": 1.2775268774229647, "grad_norm": 0.1663319617509842, "learning_rate": 7.015194236580998e-05, "loss": 0.0311, "step": 63930 }, { "epoch": 1.2777267095639662, "grad_norm": 0.15616200864315033, "learning_rate": 7.014248426510646e-05, "loss": 0.0278, "step": 63940 }, { "epoch": 1.2779265417049679, "grad_norm": 0.1600475311279297, "learning_rate": 7.013302530395303e-05, "loss": 0.039, "step": 63950 }, { "epoch": 1.2781263738459694, "grad_norm": 0.14054258167743683, "learning_rate": 7.012356548275371e-05, "loss": 0.0349, "step": 63960 }, { "epoch": 1.2783262059869709, "grad_norm": 0.13537172973155975, "learning_rate": 7.011410480191267e-05, "loss": 0.0307, "step": 63970 }, { "epoch": 1.2785260381279726, "grad_norm": 0.1239960566163063, "learning_rate": 7.010464326183401e-05, "loss": 0.0298, "step": 63980 }, { "epoch": 1.278725870268974, "grad_norm": 0.13918475806713104, "learning_rate": 7.009518086292194e-05, "loss": 0.0225, "step": 63990 }, { "epoch": 1.2789257024099756, "grad_norm": 0.13151070475578308, "learning_rate": 7.008571760558065e-05, "loss": 0.0409, "step": 64000 }, { "epoch": 1.279125534550977, "grad_norm": 0.13791586458683014, "learning_rate": 7.00762534902144e-05, "loss": 0.0274, "step": 64010 }, { "epoch": 1.2793253666919786, "grad_norm": 0.14625036716461182, "learning_rate": 7.006678851722748e-05, "loss": 0.0315, "step": 64020 }, { "epoch": 1.2795251988329803, "grad_norm": 0.15034963190555573, "learning_rate": 7.005732268702423e-05, "loss": 0.0305, "step": 64030 }, { "epoch": 1.2797250309739818, "grad_norm": 0.13905541598796844, "learning_rate": 7.0047856000009e-05, "loss": 0.034, "step": 64040 }, { "epoch": 1.2799248631149833, "grad_norm": 0.18296293914318085, "learning_rate": 7.003838845658618e-05, "loss": 0.0423, "step": 64050 }, { "epoch": 1.280124695255985, "grad_norm": 0.11885379999876022, "learning_rate": 7.002892005716023e-05, "loss": 0.0313, "step": 64060 }, { "epoch": 1.2803245273969865, "grad_norm": 0.1558949053287506, "learning_rate": 7.001945080213559e-05, "loss": 0.0294, "step": 64070 }, { "epoch": 1.280524359537988, "grad_norm": 0.11477836966514587, "learning_rate": 7.000998069191681e-05, "loss": 0.0371, "step": 64080 }, { "epoch": 1.2807241916789898, "grad_norm": 0.1439322680234909, "learning_rate": 7.00005097269084e-05, "loss": 0.0287, "step": 64090 }, { "epoch": 1.2809240238199913, "grad_norm": 0.1400790959596634, "learning_rate": 6.999103790751496e-05, "loss": 0.0278, "step": 64100 }, { "epoch": 1.2811238559609928, "grad_norm": 0.11529596149921417, "learning_rate": 6.998156523414111e-05, "loss": 0.0318, "step": 64110 }, { "epoch": 1.2813236881019943, "grad_norm": 0.0906318947672844, "learning_rate": 6.997209170719148e-05, "loss": 0.0308, "step": 64120 }, { "epoch": 1.2815235202429958, "grad_norm": 0.11895645409822464, "learning_rate": 6.996261732707082e-05, "loss": 0.0293, "step": 64130 }, { "epoch": 1.2817233523839975, "grad_norm": 0.11337077617645264, "learning_rate": 6.995314209418379e-05, "loss": 0.0281, "step": 64140 }, { "epoch": 1.281923184524999, "grad_norm": 0.11598671227693558, "learning_rate": 6.994366600893521e-05, "loss": 0.0409, "step": 64150 }, { "epoch": 1.2821230166660005, "grad_norm": 0.11682426929473877, "learning_rate": 6.993418907172983e-05, "loss": 0.0334, "step": 64160 }, { "epoch": 1.2823228488070022, "grad_norm": 0.11821211129426956, "learning_rate": 6.992471128297252e-05, "loss": 0.0309, "step": 64170 }, { "epoch": 1.2825226809480037, "grad_norm": 0.130132257938385, "learning_rate": 6.991523264306815e-05, "loss": 0.0391, "step": 64180 }, { "epoch": 1.2827225130890052, "grad_norm": 0.1274981051683426, "learning_rate": 6.990575315242163e-05, "loss": 0.0299, "step": 64190 }, { "epoch": 1.282922345230007, "grad_norm": 0.12122385948896408, "learning_rate": 6.989627281143789e-05, "loss": 0.0347, "step": 64200 }, { "epoch": 1.2831221773710084, "grad_norm": 0.13190323114395142, "learning_rate": 6.988679162052196e-05, "loss": 0.034, "step": 64210 }, { "epoch": 1.28332200951201, "grad_norm": 0.155740424990654, "learning_rate": 6.987730958007878e-05, "loss": 0.0331, "step": 64220 }, { "epoch": 1.2835218416530114, "grad_norm": 0.13357074558734894, "learning_rate": 6.986782669051348e-05, "loss": 0.0318, "step": 64230 }, { "epoch": 1.283721673794013, "grad_norm": 0.15245990455150604, "learning_rate": 6.985834295223112e-05, "loss": 0.0435, "step": 64240 }, { "epoch": 1.2839215059350146, "grad_norm": 0.14354640245437622, "learning_rate": 6.984885836563681e-05, "loss": 0.026, "step": 64250 }, { "epoch": 1.2841213380760161, "grad_norm": 0.13248707354068756, "learning_rate": 6.983937293113575e-05, "loss": 0.032, "step": 64260 }, { "epoch": 1.2843211702170176, "grad_norm": 0.1592242419719696, "learning_rate": 6.982988664913313e-05, "loss": 0.0369, "step": 64270 }, { "epoch": 1.2845210023580194, "grad_norm": 0.14035643637180328, "learning_rate": 6.982039952003418e-05, "loss": 0.0332, "step": 64280 }, { "epoch": 1.2847208344990209, "grad_norm": 0.11892440170049667, "learning_rate": 6.981091154424415e-05, "loss": 0.034, "step": 64290 }, { "epoch": 1.2849206666400224, "grad_norm": 0.16409681737422943, "learning_rate": 6.98014227221684e-05, "loss": 0.0319, "step": 64300 }, { "epoch": 1.2851204987810239, "grad_norm": 0.12838049232959747, "learning_rate": 6.979193305421223e-05, "loss": 0.0376, "step": 64310 }, { "epoch": 1.2853203309220256, "grad_norm": 0.11025632917881012, "learning_rate": 6.978244254078106e-05, "loss": 0.0335, "step": 64320 }, { "epoch": 1.285520163063027, "grad_norm": 0.15386010706424713, "learning_rate": 6.977295118228027e-05, "loss": 0.0372, "step": 64330 }, { "epoch": 1.2857199952040286, "grad_norm": 0.12650825083255768, "learning_rate": 6.976345897911532e-05, "loss": 0.0383, "step": 64340 }, { "epoch": 1.28591982734503, "grad_norm": 0.13945171236991882, "learning_rate": 6.975396593169174e-05, "loss": 0.0385, "step": 64350 }, { "epoch": 1.2861196594860318, "grad_norm": 0.1264079511165619, "learning_rate": 6.9744472040415e-05, "loss": 0.0368, "step": 64360 }, { "epoch": 1.2863194916270333, "grad_norm": 0.11206310987472534, "learning_rate": 6.97349773056907e-05, "loss": 0.0278, "step": 64370 }, { "epoch": 1.2865193237680348, "grad_norm": 0.1306075006723404, "learning_rate": 6.972548172792441e-05, "loss": 0.0261, "step": 64380 }, { "epoch": 1.2867191559090365, "grad_norm": 0.14284823834896088, "learning_rate": 6.971598530752178e-05, "loss": 0.0268, "step": 64390 }, { "epoch": 1.286918988050038, "grad_norm": 0.10122045129537582, "learning_rate": 6.970648804488848e-05, "loss": 0.0262, "step": 64400 }, { "epoch": 1.2871188201910395, "grad_norm": 0.10486333817243576, "learning_rate": 6.969698994043023e-05, "loss": 0.0353, "step": 64410 }, { "epoch": 1.287318652332041, "grad_norm": 0.184771329164505, "learning_rate": 6.968749099455273e-05, "loss": 0.0291, "step": 64420 }, { "epoch": 1.2875184844730425, "grad_norm": 0.15231983363628387, "learning_rate": 6.967799120766178e-05, "loss": 0.0324, "step": 64430 }, { "epoch": 1.2877183166140442, "grad_norm": 0.13016539812088013, "learning_rate": 6.96684905801632e-05, "loss": 0.0291, "step": 64440 }, { "epoch": 1.2879181487550457, "grad_norm": 0.13972219824790955, "learning_rate": 6.965898911246285e-05, "loss": 0.0321, "step": 64450 }, { "epoch": 1.2881179808960472, "grad_norm": 0.1456504762172699, "learning_rate": 6.964948680496659e-05, "loss": 0.0335, "step": 64460 }, { "epoch": 1.288317813037049, "grad_norm": 0.13903461396694183, "learning_rate": 6.963998365808034e-05, "loss": 0.0319, "step": 64470 }, { "epoch": 1.2885176451780505, "grad_norm": 0.15425503253936768, "learning_rate": 6.96304796722101e-05, "loss": 0.0317, "step": 64480 }, { "epoch": 1.288717477319052, "grad_norm": 0.15071457624435425, "learning_rate": 6.962097484776179e-05, "loss": 0.0381, "step": 64490 }, { "epoch": 1.2889173094600537, "grad_norm": 0.16611354053020477, "learning_rate": 6.96114691851415e-05, "loss": 0.027, "step": 64500 }, { "epoch": 1.2891171416010552, "grad_norm": 0.11390437185764313, "learning_rate": 6.960196268475528e-05, "loss": 0.0254, "step": 64510 }, { "epoch": 1.2893169737420567, "grad_norm": 0.12369677424430847, "learning_rate": 6.959245534700921e-05, "loss": 0.0302, "step": 64520 }, { "epoch": 1.2895168058830582, "grad_norm": 0.16443875432014465, "learning_rate": 6.958294717230946e-05, "loss": 0.0307, "step": 64530 }, { "epoch": 1.2897166380240597, "grad_norm": 0.09435844421386719, "learning_rate": 6.957343816106215e-05, "loss": 0.0322, "step": 64540 }, { "epoch": 1.2899164701650614, "grad_norm": 0.11764310300350189, "learning_rate": 6.956392831367354e-05, "loss": 0.0334, "step": 64550 }, { "epoch": 1.290116302306063, "grad_norm": 0.12998925149440765, "learning_rate": 6.955441763054986e-05, "loss": 0.0342, "step": 64560 }, { "epoch": 1.2903161344470644, "grad_norm": 0.1343454271554947, "learning_rate": 6.954490611209737e-05, "loss": 0.0351, "step": 64570 }, { "epoch": 1.290515966588066, "grad_norm": 0.12971164286136627, "learning_rate": 6.95353937587224e-05, "loss": 0.0341, "step": 64580 }, { "epoch": 1.2907157987290676, "grad_norm": 0.12567013502120972, "learning_rate": 6.952588057083128e-05, "loss": 0.0243, "step": 64590 }, { "epoch": 1.290915630870069, "grad_norm": 0.10607115179300308, "learning_rate": 6.951636654883042e-05, "loss": 0.0346, "step": 64600 }, { "epoch": 1.2911154630110708, "grad_norm": 0.13334961235523224, "learning_rate": 6.950685169312626e-05, "loss": 0.0378, "step": 64610 }, { "epoch": 1.2913152951520723, "grad_norm": 0.15258482098579407, "learning_rate": 6.94973360041252e-05, "loss": 0.0309, "step": 64620 }, { "epoch": 1.2915151272930738, "grad_norm": 0.11970769613981247, "learning_rate": 6.948781948223377e-05, "loss": 0.0339, "step": 64630 }, { "epoch": 1.2917149594340753, "grad_norm": 0.14052151143550873, "learning_rate": 6.947830212785851e-05, "loss": 0.028, "step": 64640 }, { "epoch": 1.2919147915750768, "grad_norm": 0.14506058394908905, "learning_rate": 6.946878394140595e-05, "loss": 0.0347, "step": 64650 }, { "epoch": 1.2921146237160785, "grad_norm": 0.12183867394924164, "learning_rate": 6.94592649232827e-05, "loss": 0.0315, "step": 64660 }, { "epoch": 1.29231445585708, "grad_norm": 0.160134956240654, "learning_rate": 6.94497450738954e-05, "loss": 0.0362, "step": 64670 }, { "epoch": 1.2925142879980815, "grad_norm": 0.14513157308101654, "learning_rate": 6.944022439365074e-05, "loss": 0.036, "step": 64680 }, { "epoch": 1.2927141201390833, "grad_norm": 0.1267022341489792, "learning_rate": 6.943070288295539e-05, "loss": 0.0334, "step": 64690 }, { "epoch": 1.2929139522800848, "grad_norm": 0.12582828104496002, "learning_rate": 6.942118054221612e-05, "loss": 0.0303, "step": 64700 }, { "epoch": 1.2931137844210863, "grad_norm": 0.14159847795963287, "learning_rate": 6.941165737183968e-05, "loss": 0.0263, "step": 64710 }, { "epoch": 1.2933136165620878, "grad_norm": 0.12768018245697021, "learning_rate": 6.940213337223292e-05, "loss": 0.0464, "step": 64720 }, { "epoch": 1.2935134487030895, "grad_norm": 0.12116550654172897, "learning_rate": 6.939260854380265e-05, "loss": 0.033, "step": 64730 }, { "epoch": 1.293713280844091, "grad_norm": 0.1381838172674179, "learning_rate": 6.938308288695576e-05, "loss": 0.0398, "step": 64740 }, { "epoch": 1.2939131129850925, "grad_norm": 0.12731656432151794, "learning_rate": 6.937355640209919e-05, "loss": 0.0311, "step": 64750 }, { "epoch": 1.294112945126094, "grad_norm": 0.129765123128891, "learning_rate": 6.936402908963989e-05, "loss": 0.0367, "step": 64760 }, { "epoch": 1.2943127772670957, "grad_norm": 0.11770352721214294, "learning_rate": 6.935450094998481e-05, "loss": 0.0415, "step": 64770 }, { "epoch": 1.2945126094080972, "grad_norm": 0.13230805099010468, "learning_rate": 6.934497198354102e-05, "loss": 0.0376, "step": 64780 }, { "epoch": 1.2947124415490987, "grad_norm": 0.13587352633476257, "learning_rate": 6.933544219071558e-05, "loss": 0.0314, "step": 64790 }, { "epoch": 1.2949122736901004, "grad_norm": 0.1370389610528946, "learning_rate": 6.932591157191554e-05, "loss": 0.0346, "step": 64800 }, { "epoch": 1.295112105831102, "grad_norm": 0.15043137967586517, "learning_rate": 6.931638012754809e-05, "loss": 0.0348, "step": 64810 }, { "epoch": 1.2953119379721034, "grad_norm": 0.14649654924869537, "learning_rate": 6.930684785802035e-05, "loss": 0.0321, "step": 64820 }, { "epoch": 1.295511770113105, "grad_norm": 0.15850557386875153, "learning_rate": 6.929731476373954e-05, "loss": 0.0369, "step": 64830 }, { "epoch": 1.2957116022541064, "grad_norm": 0.12388642877340317, "learning_rate": 6.92877808451129e-05, "loss": 0.0366, "step": 64840 }, { "epoch": 1.2959114343951081, "grad_norm": 0.1261180192232132, "learning_rate": 6.927824610254768e-05, "loss": 0.0366, "step": 64850 }, { "epoch": 1.2961112665361096, "grad_norm": 0.12569545209407806, "learning_rate": 6.926871053645122e-05, "loss": 0.0321, "step": 64860 }, { "epoch": 1.2963110986771111, "grad_norm": 0.19130532443523407, "learning_rate": 6.925917414723082e-05, "loss": 0.0376, "step": 64870 }, { "epoch": 1.2965109308181129, "grad_norm": 0.14722774922847748, "learning_rate": 6.924963693529391e-05, "loss": 0.0395, "step": 64880 }, { "epoch": 1.2967107629591144, "grad_norm": 0.14479030668735504, "learning_rate": 6.924009890104785e-05, "loss": 0.0282, "step": 64890 }, { "epoch": 1.2969105951001159, "grad_norm": 0.16818824410438538, "learning_rate": 6.923056004490012e-05, "loss": 0.0323, "step": 64900 }, { "epoch": 1.2971104272411176, "grad_norm": 0.11058970540761948, "learning_rate": 6.92210203672582e-05, "loss": 0.035, "step": 64910 }, { "epoch": 1.297310259382119, "grad_norm": 0.12615780532360077, "learning_rate": 6.921147986852958e-05, "loss": 0.027, "step": 64920 }, { "epoch": 1.2975100915231206, "grad_norm": 0.15763521194458008, "learning_rate": 6.920193854912185e-05, "loss": 0.0345, "step": 64930 }, { "epoch": 1.297709923664122, "grad_norm": 0.1287427693605423, "learning_rate": 6.919239640944257e-05, "loss": 0.0294, "step": 64940 }, { "epoch": 1.2979097558051236, "grad_norm": 0.09599866718053818, "learning_rate": 6.918285344989939e-05, "loss": 0.0242, "step": 64950 }, { "epoch": 1.2981095879461253, "grad_norm": 0.10948648303747177, "learning_rate": 6.917330967089992e-05, "loss": 0.0264, "step": 64960 }, { "epoch": 1.2983094200871268, "grad_norm": 0.09119123965501785, "learning_rate": 6.91637650728519e-05, "loss": 0.0312, "step": 64970 }, { "epoch": 1.2985092522281283, "grad_norm": 0.12467892467975616, "learning_rate": 6.915421965616306e-05, "loss": 0.039, "step": 64980 }, { "epoch": 1.29870908436913, "grad_norm": 0.13673758506774902, "learning_rate": 6.914467342124111e-05, "loss": 0.0358, "step": 64990 }, { "epoch": 1.2989089165101315, "grad_norm": 0.1326395869255066, "learning_rate": 6.913512636849391e-05, "loss": 0.0317, "step": 65000 }, { "epoch": 1.299108748651133, "grad_norm": 0.1289660632610321, "learning_rate": 6.912557849832925e-05, "loss": 0.0351, "step": 65010 }, { "epoch": 1.2993085807921347, "grad_norm": 0.1355704367160797, "learning_rate": 6.911602981115502e-05, "loss": 0.0388, "step": 65020 }, { "epoch": 1.2995084129331362, "grad_norm": 0.14314807951450348, "learning_rate": 6.91064803073791e-05, "loss": 0.0319, "step": 65030 }, { "epoch": 1.2997082450741377, "grad_norm": 0.12908846139907837, "learning_rate": 6.909692998740945e-05, "loss": 0.0301, "step": 65040 }, { "epoch": 1.2999080772151392, "grad_norm": 0.14011569321155548, "learning_rate": 6.908737885165401e-05, "loss": 0.03, "step": 65050 }, { "epoch": 1.3001079093561407, "grad_norm": 0.14639371633529663, "learning_rate": 6.907782690052085e-05, "loss": 0.034, "step": 65060 }, { "epoch": 1.3003077414971425, "grad_norm": 0.13858158886432648, "learning_rate": 6.906827413441795e-05, "loss": 0.0397, "step": 65070 }, { "epoch": 1.300507573638144, "grad_norm": 0.1366656869649887, "learning_rate": 6.905872055375341e-05, "loss": 0.0398, "step": 65080 }, { "epoch": 1.3007074057791455, "grad_norm": 0.11942554265260696, "learning_rate": 6.904916615893533e-05, "loss": 0.0302, "step": 65090 }, { "epoch": 1.3009072379201472, "grad_norm": 0.16975192725658417, "learning_rate": 6.903961095037188e-05, "loss": 0.038, "step": 65100 }, { "epoch": 1.3011070700611487, "grad_norm": 0.16335073113441467, "learning_rate": 6.903005492847122e-05, "loss": 0.0352, "step": 65110 }, { "epoch": 1.3013069022021502, "grad_norm": 0.15783481299877167, "learning_rate": 6.902049809364157e-05, "loss": 0.0328, "step": 65120 }, { "epoch": 1.3015067343431517, "grad_norm": 0.13493074476718903, "learning_rate": 6.90109404462912e-05, "loss": 0.032, "step": 65130 }, { "epoch": 1.3017065664841534, "grad_norm": 0.1857634037733078, "learning_rate": 6.900138198682836e-05, "loss": 0.0415, "step": 65140 }, { "epoch": 1.301906398625155, "grad_norm": 0.13227559626102448, "learning_rate": 6.89918227156614e-05, "loss": 0.0418, "step": 65150 }, { "epoch": 1.3021062307661564, "grad_norm": 0.17955195903778076, "learning_rate": 6.898226263319867e-05, "loss": 0.0419, "step": 65160 }, { "epoch": 1.302306062907158, "grad_norm": 0.13311056792736053, "learning_rate": 6.897270173984854e-05, "loss": 0.0335, "step": 65170 }, { "epoch": 1.3025058950481596, "grad_norm": 0.11526482552289963, "learning_rate": 6.896314003601945e-05, "loss": 0.0321, "step": 65180 }, { "epoch": 1.3027057271891611, "grad_norm": 0.10163382440805435, "learning_rate": 6.895357752211984e-05, "loss": 0.0434, "step": 65190 }, { "epoch": 1.3029055593301626, "grad_norm": 0.12282057106494904, "learning_rate": 6.894401419855823e-05, "loss": 0.0326, "step": 65200 }, { "epoch": 1.3031053914711643, "grad_norm": 0.10494188964366913, "learning_rate": 6.893445006574314e-05, "loss": 0.0324, "step": 65210 }, { "epoch": 1.3033052236121658, "grad_norm": 0.1450147032737732, "learning_rate": 6.892488512408311e-05, "loss": 0.0262, "step": 65220 }, { "epoch": 1.3035050557531673, "grad_norm": 0.1348714977502823, "learning_rate": 6.891531937398677e-05, "loss": 0.0342, "step": 65230 }, { "epoch": 1.3037048878941688, "grad_norm": 0.12529470026493073, "learning_rate": 6.890575281586273e-05, "loss": 0.0319, "step": 65240 }, { "epoch": 1.3039047200351703, "grad_norm": 0.13573725521564484, "learning_rate": 6.889618545011964e-05, "loss": 0.0295, "step": 65250 }, { "epoch": 1.304104552176172, "grad_norm": 0.14496557414531708, "learning_rate": 6.888661727716625e-05, "loss": 0.0286, "step": 65260 }, { "epoch": 1.3043043843171735, "grad_norm": 0.12085206061601639, "learning_rate": 6.887704829741126e-05, "loss": 0.0264, "step": 65270 }, { "epoch": 1.304504216458175, "grad_norm": 0.14613275229930878, "learning_rate": 6.886747851126344e-05, "loss": 0.0396, "step": 65280 }, { "epoch": 1.3047040485991768, "grad_norm": 0.13871294260025024, "learning_rate": 6.885790791913161e-05, "loss": 0.0361, "step": 65290 }, { "epoch": 1.3049038807401783, "grad_norm": 0.14191821217536926, "learning_rate": 6.884833652142457e-05, "loss": 0.0367, "step": 65300 }, { "epoch": 1.3051037128811798, "grad_norm": 0.14022405445575714, "learning_rate": 6.883876431855123e-05, "loss": 0.0306, "step": 65310 }, { "epoch": 1.3053035450221815, "grad_norm": 0.12486207485198975, "learning_rate": 6.882919131092049e-05, "loss": 0.0237, "step": 65320 }, { "epoch": 1.305503377163183, "grad_norm": 0.14177510142326355, "learning_rate": 6.88196174989413e-05, "loss": 0.0313, "step": 65330 }, { "epoch": 1.3057032093041845, "grad_norm": 0.19032502174377441, "learning_rate": 6.88100428830226e-05, "loss": 0.0434, "step": 65340 }, { "epoch": 1.305903041445186, "grad_norm": 0.13007627427577972, "learning_rate": 6.880046746357344e-05, "loss": 0.0333, "step": 65350 }, { "epoch": 1.3061028735861875, "grad_norm": 0.15846532583236694, "learning_rate": 6.879089124100284e-05, "loss": 0.0343, "step": 65360 }, { "epoch": 1.3063027057271892, "grad_norm": 0.13279321789741516, "learning_rate": 6.87813142157199e-05, "loss": 0.0262, "step": 65370 }, { "epoch": 1.3065025378681907, "grad_norm": 0.1275872439146042, "learning_rate": 6.877173638813371e-05, "loss": 0.03, "step": 65380 }, { "epoch": 1.3067023700091922, "grad_norm": 0.15816357731819153, "learning_rate": 6.876215775865341e-05, "loss": 0.032, "step": 65390 }, { "epoch": 1.306902202150194, "grad_norm": 0.11715829372406006, "learning_rate": 6.875257832768822e-05, "loss": 0.0327, "step": 65400 }, { "epoch": 1.3071020342911954, "grad_norm": 0.11784087121486664, "learning_rate": 6.874299809564732e-05, "loss": 0.0365, "step": 65410 }, { "epoch": 1.307301866432197, "grad_norm": 0.14615346491336823, "learning_rate": 6.873341706293999e-05, "loss": 0.0436, "step": 65420 }, { "epoch": 1.3075016985731986, "grad_norm": 0.10123865306377411, "learning_rate": 6.872383522997547e-05, "loss": 0.0265, "step": 65430 }, { "epoch": 1.3077015307142001, "grad_norm": 0.11921711266040802, "learning_rate": 6.871425259716314e-05, "loss": 0.0362, "step": 65440 }, { "epoch": 1.3079013628552016, "grad_norm": 0.13123932480812073, "learning_rate": 6.87046691649123e-05, "loss": 0.0287, "step": 65450 }, { "epoch": 1.3081011949962031, "grad_norm": 0.15741218626499176, "learning_rate": 6.869508493363234e-05, "loss": 0.0255, "step": 65460 }, { "epoch": 1.3083010271372046, "grad_norm": 0.10996469110250473, "learning_rate": 6.868549990373272e-05, "loss": 0.0314, "step": 65470 }, { "epoch": 1.3085008592782064, "grad_norm": 0.10917437076568604, "learning_rate": 6.867591407562287e-05, "loss": 0.0318, "step": 65480 }, { "epoch": 1.3087006914192079, "grad_norm": 0.12290076166391373, "learning_rate": 6.866632744971227e-05, "loss": 0.0329, "step": 65490 }, { "epoch": 1.3089005235602094, "grad_norm": 0.11787158250808716, "learning_rate": 6.865674002641043e-05, "loss": 0.0344, "step": 65500 }, { "epoch": 1.309100355701211, "grad_norm": 0.13168564438819885, "learning_rate": 6.864715180612696e-05, "loss": 0.0285, "step": 65510 }, { "epoch": 1.3093001878422126, "grad_norm": 0.11698441207408905, "learning_rate": 6.863756278927142e-05, "loss": 0.0323, "step": 65520 }, { "epoch": 1.309500019983214, "grad_norm": 0.1303323358297348, "learning_rate": 6.862797297625343e-05, "loss": 0.0339, "step": 65530 }, { "epoch": 1.3096998521242156, "grad_norm": 0.13055956363677979, "learning_rate": 6.861838236748266e-05, "loss": 0.0407, "step": 65540 }, { "epoch": 1.309899684265217, "grad_norm": 0.14362025260925293, "learning_rate": 6.860879096336878e-05, "loss": 0.0306, "step": 65550 }, { "epoch": 1.3100995164062188, "grad_norm": 0.16485024988651276, "learning_rate": 6.859919876432154e-05, "loss": 0.0333, "step": 65560 }, { "epoch": 1.3102993485472203, "grad_norm": 0.13933035731315613, "learning_rate": 6.85896057707507e-05, "loss": 0.0324, "step": 65570 }, { "epoch": 1.3104991806882218, "grad_norm": 0.10768365859985352, "learning_rate": 6.858001198306608e-05, "loss": 0.0261, "step": 65580 }, { "epoch": 1.3106990128292235, "grad_norm": 0.11459843814373016, "learning_rate": 6.857041740167745e-05, "loss": 0.0288, "step": 65590 }, { "epoch": 1.310898844970225, "grad_norm": 0.11286374181509018, "learning_rate": 6.856082202699473e-05, "loss": 0.0382, "step": 65600 }, { "epoch": 1.3110986771112265, "grad_norm": 0.13549214601516724, "learning_rate": 6.855122585942775e-05, "loss": 0.0383, "step": 65610 }, { "epoch": 1.3112985092522282, "grad_norm": 0.16802850365638733, "learning_rate": 6.85416288993865e-05, "loss": 0.0355, "step": 65620 }, { "epoch": 1.3114983413932297, "grad_norm": 0.09552840143442154, "learning_rate": 6.853203114728094e-05, "loss": 0.0255, "step": 65630 }, { "epoch": 1.3116981735342312, "grad_norm": 0.18277917802333832, "learning_rate": 6.852243260352103e-05, "loss": 0.0304, "step": 65640 }, { "epoch": 1.3118980056752327, "grad_norm": 0.12360184639692307, "learning_rate": 6.851283326851685e-05, "loss": 0.0311, "step": 65650 }, { "epoch": 1.3120978378162342, "grad_norm": 0.10060800611972809, "learning_rate": 6.850323314267843e-05, "loss": 0.029, "step": 65660 }, { "epoch": 1.312297669957236, "grad_norm": 0.10403578728437424, "learning_rate": 6.849363222641587e-05, "loss": 0.041, "step": 65670 }, { "epoch": 1.3124975020982375, "grad_norm": 0.16323687136173248, "learning_rate": 6.848403052013933e-05, "loss": 0.0282, "step": 65680 }, { "epoch": 1.312697334239239, "grad_norm": 0.15483349561691284, "learning_rate": 6.847442802425895e-05, "loss": 0.0409, "step": 65690 }, { "epoch": 1.3128971663802407, "grad_norm": 0.1326247900724411, "learning_rate": 6.846482473918495e-05, "loss": 0.0291, "step": 65700 }, { "epoch": 1.3130969985212422, "grad_norm": 0.11978331208229065, "learning_rate": 6.845522066532757e-05, "loss": 0.0271, "step": 65710 }, { "epoch": 1.3132968306622437, "grad_norm": 0.14504310488700867, "learning_rate": 6.844561580309705e-05, "loss": 0.0326, "step": 65720 }, { "epoch": 1.3134966628032454, "grad_norm": 0.13701856136322021, "learning_rate": 6.843601015290371e-05, "loss": 0.0337, "step": 65730 }, { "epoch": 1.313696494944247, "grad_norm": 0.11105488240718842, "learning_rate": 6.84264037151579e-05, "loss": 0.0293, "step": 65740 }, { "epoch": 1.3138963270852484, "grad_norm": 0.13978934288024902, "learning_rate": 6.841679649026997e-05, "loss": 0.0275, "step": 65750 }, { "epoch": 1.31409615922625, "grad_norm": 0.1260833740234375, "learning_rate": 6.840718847865031e-05, "loss": 0.0354, "step": 65760 }, { "epoch": 1.3142959913672514, "grad_norm": 0.1385996788740158, "learning_rate": 6.839757968070937e-05, "loss": 0.0258, "step": 65770 }, { "epoch": 1.3144958235082531, "grad_norm": 0.11655033379793167, "learning_rate": 6.838797009685766e-05, "loss": 0.0392, "step": 65780 }, { "epoch": 1.3146956556492546, "grad_norm": 0.12662862241268158, "learning_rate": 6.837835972750561e-05, "loss": 0.0319, "step": 65790 }, { "epoch": 1.3148954877902561, "grad_norm": 0.1435818374156952, "learning_rate": 6.836874857306381e-05, "loss": 0.0373, "step": 65800 }, { "epoch": 1.3150953199312578, "grad_norm": 0.14885565638542175, "learning_rate": 6.835913663394282e-05, "loss": 0.0292, "step": 65810 }, { "epoch": 1.3152951520722593, "grad_norm": 0.10517649352550507, "learning_rate": 6.834952391055322e-05, "loss": 0.0318, "step": 65820 }, { "epoch": 1.3154949842132608, "grad_norm": 0.14220120012760162, "learning_rate": 6.833991040330569e-05, "loss": 0.0292, "step": 65830 }, { "epoch": 1.3156948163542626, "grad_norm": 0.12239594012498856, "learning_rate": 6.833029611261088e-05, "loss": 0.0247, "step": 65840 }, { "epoch": 1.315894648495264, "grad_norm": 0.13988657295703888, "learning_rate": 6.832068103887948e-05, "loss": 0.0295, "step": 65850 }, { "epoch": 1.3160944806362656, "grad_norm": 0.12269330024719238, "learning_rate": 6.831106518252224e-05, "loss": 0.0281, "step": 65860 }, { "epoch": 1.316294312777267, "grad_norm": 0.1376120150089264, "learning_rate": 6.830144854394994e-05, "loss": 0.0323, "step": 65870 }, { "epoch": 1.3164941449182685, "grad_norm": 0.10967165231704712, "learning_rate": 6.829183112357338e-05, "loss": 0.0314, "step": 65880 }, { "epoch": 1.3166939770592703, "grad_norm": 0.1343129426240921, "learning_rate": 6.828221292180339e-05, "loss": 0.0311, "step": 65890 }, { "epoch": 1.3168938092002718, "grad_norm": 0.13306564092636108, "learning_rate": 6.827259393905085e-05, "loss": 0.0389, "step": 65900 }, { "epoch": 1.3170936413412733, "grad_norm": 0.13254275918006897, "learning_rate": 6.826297417572668e-05, "loss": 0.0344, "step": 65910 }, { "epoch": 1.317293473482275, "grad_norm": 0.12237668037414551, "learning_rate": 6.82533536322418e-05, "loss": 0.034, "step": 65920 }, { "epoch": 1.3174933056232765, "grad_norm": 0.1490425169467926, "learning_rate": 6.824373230900718e-05, "loss": 0.0413, "step": 65930 }, { "epoch": 1.317693137764278, "grad_norm": 0.15406429767608643, "learning_rate": 6.823411020643383e-05, "loss": 0.0326, "step": 65940 }, { "epoch": 1.3178929699052795, "grad_norm": 0.1527700126171112, "learning_rate": 6.822448732493278e-05, "loss": 0.0324, "step": 65950 }, { "epoch": 1.318092802046281, "grad_norm": 0.15034495294094086, "learning_rate": 6.821486366491513e-05, "loss": 0.0399, "step": 65960 }, { "epoch": 1.3182926341872827, "grad_norm": 0.14084690809249878, "learning_rate": 6.820523922679196e-05, "loss": 0.0262, "step": 65970 }, { "epoch": 1.3184924663282842, "grad_norm": 0.11461113393306732, "learning_rate": 6.819561401097441e-05, "loss": 0.0262, "step": 65980 }, { "epoch": 1.3186922984692857, "grad_norm": 0.11044413596391678, "learning_rate": 6.818598801787365e-05, "loss": 0.024, "step": 65990 }, { "epoch": 1.3188921306102874, "grad_norm": 0.11376601457595825, "learning_rate": 6.81763612479009e-05, "loss": 0.0347, "step": 66000 }, { "epoch": 1.319091962751289, "grad_norm": 0.12297973036766052, "learning_rate": 6.816673370146739e-05, "loss": 0.0317, "step": 66010 }, { "epoch": 1.3192917948922904, "grad_norm": 0.11469054222106934, "learning_rate": 6.815710537898438e-05, "loss": 0.031, "step": 66020 }, { "epoch": 1.3194916270332921, "grad_norm": 0.1651078462600708, "learning_rate": 6.81474762808632e-05, "loss": 0.0248, "step": 66030 }, { "epoch": 1.3196914591742936, "grad_norm": 0.13285253942012787, "learning_rate": 6.813784640751515e-05, "loss": 0.0343, "step": 66040 }, { "epoch": 1.3198912913152951, "grad_norm": 0.1290573626756668, "learning_rate": 6.812821575935164e-05, "loss": 0.0347, "step": 66050 }, { "epoch": 1.3200911234562966, "grad_norm": 0.12868495285511017, "learning_rate": 6.811858433678404e-05, "loss": 0.0266, "step": 66060 }, { "epoch": 1.3202909555972981, "grad_norm": 0.1391085982322693, "learning_rate": 6.810895214022381e-05, "loss": 0.0352, "step": 66070 }, { "epoch": 1.3204907877382999, "grad_norm": 0.17149588465690613, "learning_rate": 6.809931917008241e-05, "loss": 0.0279, "step": 66080 }, { "epoch": 1.3206906198793014, "grad_norm": 0.1306273639202118, "learning_rate": 6.808968542677135e-05, "loss": 0.0315, "step": 66090 }, { "epoch": 1.3208904520203029, "grad_norm": 0.11656003445386887, "learning_rate": 6.808005091070217e-05, "loss": 0.0333, "step": 66100 }, { "epoch": 1.3210902841613046, "grad_norm": 0.11514105647802353, "learning_rate": 6.80704156222864e-05, "loss": 0.0315, "step": 66110 }, { "epoch": 1.321290116302306, "grad_norm": 0.1401057094335556, "learning_rate": 6.806077956193568e-05, "loss": 0.0342, "step": 66120 }, { "epoch": 1.3214899484433076, "grad_norm": 0.13894642889499664, "learning_rate": 6.805114273006166e-05, "loss": 0.0257, "step": 66130 }, { "epoch": 1.3216897805843093, "grad_norm": 0.13749772310256958, "learning_rate": 6.804150512707596e-05, "loss": 0.0334, "step": 66140 }, { "epoch": 1.3218896127253108, "grad_norm": 0.10170675069093704, "learning_rate": 6.80318667533903e-05, "loss": 0.0298, "step": 66150 }, { "epoch": 1.3220894448663123, "grad_norm": 0.12142062932252884, "learning_rate": 6.802222760941643e-05, "loss": 0.0251, "step": 66160 }, { "epoch": 1.3222892770073138, "grad_norm": 0.14792650938034058, "learning_rate": 6.801258769556609e-05, "loss": 0.0257, "step": 66170 }, { "epoch": 1.3224891091483153, "grad_norm": 0.09945005923509598, "learning_rate": 6.800294701225113e-05, "loss": 0.033, "step": 66180 }, { "epoch": 1.322688941289317, "grad_norm": 0.11982979625463486, "learning_rate": 6.799330555988333e-05, "loss": 0.0362, "step": 66190 }, { "epoch": 1.3228887734303185, "grad_norm": 0.10642972588539124, "learning_rate": 6.798366333887455e-05, "loss": 0.0323, "step": 66200 }, { "epoch": 1.32308860557132, "grad_norm": 0.15328428149223328, "learning_rate": 6.797402034963673e-05, "loss": 0.0352, "step": 66210 }, { "epoch": 1.3232884377123217, "grad_norm": 0.11222425103187561, "learning_rate": 6.796437659258177e-05, "loss": 0.0323, "step": 66220 }, { "epoch": 1.3234882698533232, "grad_norm": 0.12446131557226181, "learning_rate": 6.795473206812165e-05, "loss": 0.0309, "step": 66230 }, { "epoch": 1.3236881019943247, "grad_norm": 0.1361614465713501, "learning_rate": 6.794508677666834e-05, "loss": 0.025, "step": 66240 }, { "epoch": 1.3238879341353265, "grad_norm": 0.18456165492534637, "learning_rate": 6.793544071863392e-05, "loss": 0.0282, "step": 66250 }, { "epoch": 1.324087766276328, "grad_norm": 0.13133811950683594, "learning_rate": 6.792579389443039e-05, "loss": 0.0316, "step": 66260 }, { "epoch": 1.3242875984173295, "grad_norm": 0.14296351373195648, "learning_rate": 6.79161463044699e-05, "loss": 0.0354, "step": 66270 }, { "epoch": 1.324487430558331, "grad_norm": 0.13299275934696198, "learning_rate": 6.790649794916456e-05, "loss": 0.0336, "step": 66280 }, { "epoch": 1.3246872626993325, "grad_norm": 0.1444264054298401, "learning_rate": 6.78968488289265e-05, "loss": 0.0447, "step": 66290 }, { "epoch": 1.3248870948403342, "grad_norm": 0.14014513790607452, "learning_rate": 6.788719894416795e-05, "loss": 0.028, "step": 66300 }, { "epoch": 1.3250869269813357, "grad_norm": 0.10788396745920181, "learning_rate": 6.78775482953011e-05, "loss": 0.0325, "step": 66310 }, { "epoch": 1.3252867591223372, "grad_norm": 0.1453782469034195, "learning_rate": 6.786789688273825e-05, "loss": 0.027, "step": 66320 }, { "epoch": 1.325486591263339, "grad_norm": 0.1621972620487213, "learning_rate": 6.785824470689166e-05, "loss": 0.0375, "step": 66330 }, { "epoch": 1.3256864234043404, "grad_norm": 0.11243212968111038, "learning_rate": 6.784859176817367e-05, "loss": 0.0344, "step": 66340 }, { "epoch": 1.325886255545342, "grad_norm": 0.10351967066526413, "learning_rate": 6.783893806699662e-05, "loss": 0.0328, "step": 66350 }, { "epoch": 1.3260860876863434, "grad_norm": 0.13887815177440643, "learning_rate": 6.78292836037729e-05, "loss": 0.0308, "step": 66360 }, { "epoch": 1.326285919827345, "grad_norm": 0.16089805960655212, "learning_rate": 6.781962837891495e-05, "loss": 0.0274, "step": 66370 }, { "epoch": 1.3264857519683466, "grad_norm": 0.14416782557964325, "learning_rate": 6.780997239283523e-05, "loss": 0.0417, "step": 66380 }, { "epoch": 1.3266855841093481, "grad_norm": 0.17839136719703674, "learning_rate": 6.780031564594618e-05, "loss": 0.0431, "step": 66390 }, { "epoch": 1.3268854162503496, "grad_norm": 0.15797840058803558, "learning_rate": 6.779065813866035e-05, "loss": 0.0326, "step": 66400 }, { "epoch": 1.3270852483913513, "grad_norm": 0.1697920560836792, "learning_rate": 6.778099987139029e-05, "loss": 0.0335, "step": 66410 }, { "epoch": 1.3272850805323528, "grad_norm": 0.1320420503616333, "learning_rate": 6.777134084454857e-05, "loss": 0.0312, "step": 66420 }, { "epoch": 1.3274849126733543, "grad_norm": 0.13398270308971405, "learning_rate": 6.776168105854785e-05, "loss": 0.0338, "step": 66430 }, { "epoch": 1.327684744814356, "grad_norm": 0.19185619056224823, "learning_rate": 6.775202051380071e-05, "loss": 0.0263, "step": 66440 }, { "epoch": 1.3278845769553576, "grad_norm": 0.19445665180683136, "learning_rate": 6.774235921071987e-05, "loss": 0.0329, "step": 66450 }, { "epoch": 1.328084409096359, "grad_norm": 0.15546086430549622, "learning_rate": 6.773269714971803e-05, "loss": 0.032, "step": 66460 }, { "epoch": 1.3282842412373606, "grad_norm": 0.14877188205718994, "learning_rate": 6.772303433120798e-05, "loss": 0.0309, "step": 66470 }, { "epoch": 1.328484073378362, "grad_norm": 0.12517096102237701, "learning_rate": 6.771337075560243e-05, "loss": 0.0323, "step": 66480 }, { "epoch": 1.3286839055193638, "grad_norm": 0.12408535182476044, "learning_rate": 6.770370642331424e-05, "loss": 0.0323, "step": 66490 }, { "epoch": 1.3288837376603653, "grad_norm": 0.1274128258228302, "learning_rate": 6.769404133475623e-05, "loss": 0.0278, "step": 66500 }, { "epoch": 1.3290835698013668, "grad_norm": 0.10897555202245712, "learning_rate": 6.768437549034126e-05, "loss": 0.0318, "step": 66510 }, { "epoch": 1.3292834019423685, "grad_norm": 0.10872724652290344, "learning_rate": 6.767470889048229e-05, "loss": 0.0273, "step": 66520 }, { "epoch": 1.32948323408337, "grad_norm": 0.21155604720115662, "learning_rate": 6.76650415355922e-05, "loss": 0.0282, "step": 66530 }, { "epoch": 1.3296830662243715, "grad_norm": 0.15982313454151154, "learning_rate": 6.765537342608399e-05, "loss": 0.0429, "step": 66540 }, { "epoch": 1.3298828983653732, "grad_norm": 0.12953004240989685, "learning_rate": 6.764570456237067e-05, "loss": 0.0342, "step": 66550 }, { "epoch": 1.3300827305063747, "grad_norm": 0.13651935756206512, "learning_rate": 6.763603494486525e-05, "loss": 0.036, "step": 66560 }, { "epoch": 1.3302825626473762, "grad_norm": 0.16866734623908997, "learning_rate": 6.762636457398085e-05, "loss": 0.0281, "step": 66570 }, { "epoch": 1.3304823947883777, "grad_norm": 0.10171396285295486, "learning_rate": 6.761669345013051e-05, "loss": 0.0268, "step": 66580 }, { "epoch": 1.3306822269293792, "grad_norm": 0.10788775235414505, "learning_rate": 6.760702157372739e-05, "loss": 0.0354, "step": 66590 }, { "epoch": 1.330882059070381, "grad_norm": 0.11570718139410019, "learning_rate": 6.759734894518466e-05, "loss": 0.0351, "step": 66600 }, { "epoch": 1.3310818912113824, "grad_norm": 0.1405339539051056, "learning_rate": 6.758767556491551e-05, "loss": 0.0303, "step": 66610 }, { "epoch": 1.331281723352384, "grad_norm": 0.12829023599624634, "learning_rate": 6.757800143333316e-05, "loss": 0.0282, "step": 66620 }, { "epoch": 1.3314815554933856, "grad_norm": 0.13370153307914734, "learning_rate": 6.756832655085091e-05, "loss": 0.0383, "step": 66630 }, { "epoch": 1.3316813876343871, "grad_norm": 0.122385673224926, "learning_rate": 6.755865091788198e-05, "loss": 0.0342, "step": 66640 }, { "epoch": 1.3318812197753886, "grad_norm": 0.11237704008817673, "learning_rate": 6.754897453483977e-05, "loss": 0.0353, "step": 66650 }, { "epoch": 1.3320810519163904, "grad_norm": 0.14072902500629425, "learning_rate": 6.753929740213759e-05, "loss": 0.044, "step": 66660 }, { "epoch": 1.3322808840573919, "grad_norm": 0.12538322806358337, "learning_rate": 6.752961952018885e-05, "loss": 0.0345, "step": 66670 }, { "epoch": 1.3324807161983934, "grad_norm": 0.11269675195217133, "learning_rate": 6.751994088940697e-05, "loss": 0.0294, "step": 66680 }, { "epoch": 1.3326805483393949, "grad_norm": 0.1355884075164795, "learning_rate": 6.75102615102054e-05, "loss": 0.033, "step": 66690 }, { "epoch": 1.3328803804803964, "grad_norm": 0.10312864929437637, "learning_rate": 6.750058138299763e-05, "loss": 0.029, "step": 66700 }, { "epoch": 1.333080212621398, "grad_norm": 0.1281326562166214, "learning_rate": 6.749090050819715e-05, "loss": 0.0323, "step": 66710 }, { "epoch": 1.3332800447623996, "grad_norm": 0.10040454566478729, "learning_rate": 6.748121888621757e-05, "loss": 0.0383, "step": 66720 }, { "epoch": 1.333479876903401, "grad_norm": 0.13584308326244354, "learning_rate": 6.747153651747241e-05, "loss": 0.0347, "step": 66730 }, { "epoch": 1.3336797090444028, "grad_norm": 0.1697755753993988, "learning_rate": 6.746185340237531e-05, "loss": 0.0472, "step": 66740 }, { "epoch": 1.3338795411854043, "grad_norm": 0.17917141318321228, "learning_rate": 6.745216954133991e-05, "loss": 0.0319, "step": 66750 }, { "epoch": 1.3340793733264058, "grad_norm": 0.14282414317131042, "learning_rate": 6.744248493477988e-05, "loss": 0.0356, "step": 66760 }, { "epoch": 1.3342792054674073, "grad_norm": 0.1480533927679062, "learning_rate": 6.743279958310895e-05, "loss": 0.0297, "step": 66770 }, { "epoch": 1.3344790376084088, "grad_norm": 0.1194228082895279, "learning_rate": 6.742311348674085e-05, "loss": 0.0348, "step": 66780 }, { "epoch": 1.3346788697494105, "grad_norm": 0.10523297637701035, "learning_rate": 6.741342664608933e-05, "loss": 0.0292, "step": 66790 }, { "epoch": 1.334878701890412, "grad_norm": 0.13493621349334717, "learning_rate": 6.740373906156823e-05, "loss": 0.0332, "step": 66800 }, { "epoch": 1.3350785340314135, "grad_norm": 0.1344275325536728, "learning_rate": 6.739405073359136e-05, "loss": 0.0292, "step": 66810 }, { "epoch": 1.3352783661724152, "grad_norm": 0.12422477453947067, "learning_rate": 6.73843616625726e-05, "loss": 0.0306, "step": 66820 }, { "epoch": 1.3354781983134167, "grad_norm": 0.10841746628284454, "learning_rate": 6.737467184892585e-05, "loss": 0.0315, "step": 66830 }, { "epoch": 1.3356780304544182, "grad_norm": 0.13080528378486633, "learning_rate": 6.736498129306504e-05, "loss": 0.0337, "step": 66840 }, { "epoch": 1.33587786259542, "grad_norm": 0.11562585830688477, "learning_rate": 6.735528999540411e-05, "loss": 0.0262, "step": 66850 }, { "epoch": 1.3360776947364215, "grad_norm": 0.14347530901432037, "learning_rate": 6.734559795635709e-05, "loss": 0.0323, "step": 66860 }, { "epoch": 1.336277526877423, "grad_norm": 0.14039723575115204, "learning_rate": 6.733590517633798e-05, "loss": 0.0261, "step": 66870 }, { "epoch": 1.3364773590184245, "grad_norm": 0.14282557368278503, "learning_rate": 6.732621165576087e-05, "loss": 0.0343, "step": 66880 }, { "epoch": 1.336677191159426, "grad_norm": 0.11957409977912903, "learning_rate": 6.731651739503981e-05, "loss": 0.0223, "step": 66890 }, { "epoch": 1.3368770233004277, "grad_norm": 0.1100395917892456, "learning_rate": 6.730682239458895e-05, "loss": 0.0316, "step": 66900 }, { "epoch": 1.3370768554414292, "grad_norm": 0.158133402466774, "learning_rate": 6.72971266548224e-05, "loss": 0.0349, "step": 66910 }, { "epoch": 1.3372766875824307, "grad_norm": 0.13635818660259247, "learning_rate": 6.72874301761544e-05, "loss": 0.0354, "step": 66920 }, { "epoch": 1.3374765197234324, "grad_norm": 0.1261359006166458, "learning_rate": 6.727773295899914e-05, "loss": 0.0284, "step": 66930 }, { "epoch": 1.337676351864434, "grad_norm": 0.11763900518417358, "learning_rate": 6.726803500377087e-05, "loss": 0.0293, "step": 66940 }, { "epoch": 1.3378761840054354, "grad_norm": 0.12070353329181671, "learning_rate": 6.725833631088386e-05, "loss": 0.0304, "step": 66950 }, { "epoch": 1.3380760161464371, "grad_norm": 0.12476029247045517, "learning_rate": 6.724863688075243e-05, "loss": 0.0309, "step": 66960 }, { "epoch": 1.3382758482874386, "grad_norm": 0.12565825879573822, "learning_rate": 6.723893671379092e-05, "loss": 0.0321, "step": 66970 }, { "epoch": 1.3384756804284401, "grad_norm": 0.1402445286512375, "learning_rate": 6.72292358104137e-05, "loss": 0.031, "step": 66980 }, { "epoch": 1.3386755125694416, "grad_norm": 0.12066936492919922, "learning_rate": 6.721953417103519e-05, "loss": 0.0378, "step": 66990 }, { "epoch": 1.3388753447104431, "grad_norm": 0.13801202178001404, "learning_rate": 6.72098317960698e-05, "loss": 0.0341, "step": 67000 }, { "epoch": 1.3390751768514448, "grad_norm": 0.11484930664300919, "learning_rate": 6.7200128685932e-05, "loss": 0.0368, "step": 67010 }, { "epoch": 1.3392750089924463, "grad_norm": 0.1580851823091507, "learning_rate": 6.719042484103632e-05, "loss": 0.0349, "step": 67020 }, { "epoch": 1.3394748411334478, "grad_norm": 0.1365129053592682, "learning_rate": 6.718072026179727e-05, "loss": 0.0309, "step": 67030 }, { "epoch": 1.3396746732744496, "grad_norm": 0.13282454013824463, "learning_rate": 6.717101494862941e-05, "loss": 0.0275, "step": 67040 }, { "epoch": 1.339874505415451, "grad_norm": 0.10745050758123398, "learning_rate": 6.716130890194732e-05, "loss": 0.0361, "step": 67050 }, { "epoch": 1.3400743375564526, "grad_norm": 0.1298314779996872, "learning_rate": 6.715160212216565e-05, "loss": 0.0322, "step": 67060 }, { "epoch": 1.3402741696974543, "grad_norm": 0.1414036899805069, "learning_rate": 6.714189460969905e-05, "loss": 0.0254, "step": 67070 }, { "epoch": 1.3404740018384558, "grad_norm": 0.12092438340187073, "learning_rate": 6.71321863649622e-05, "loss": 0.0395, "step": 67080 }, { "epoch": 1.3406738339794573, "grad_norm": 0.1400415003299713, "learning_rate": 6.712247738836981e-05, "loss": 0.0336, "step": 67090 }, { "epoch": 1.3408736661204588, "grad_norm": 0.15021495521068573, "learning_rate": 6.711276768033668e-05, "loss": 0.0328, "step": 67100 }, { "epoch": 1.3410734982614603, "grad_norm": 0.12782789766788483, "learning_rate": 6.71030572412775e-05, "loss": 0.0298, "step": 67110 }, { "epoch": 1.341273330402462, "grad_norm": 0.1579630821943283, "learning_rate": 6.709334607160718e-05, "loss": 0.0338, "step": 67120 }, { "epoch": 1.3414731625434635, "grad_norm": 0.1978655904531479, "learning_rate": 6.708363417174048e-05, "loss": 0.0276, "step": 67130 }, { "epoch": 1.341672994684465, "grad_norm": 0.09353888779878616, "learning_rate": 6.707392154209234e-05, "loss": 0.0285, "step": 67140 }, { "epoch": 1.3418728268254667, "grad_norm": 0.15432493388652802, "learning_rate": 6.706420818307764e-05, "loss": 0.0373, "step": 67150 }, { "epoch": 1.3420726589664682, "grad_norm": 0.14626473188400269, "learning_rate": 6.705449409511129e-05, "loss": 0.0295, "step": 67160 }, { "epoch": 1.3422724911074697, "grad_norm": 0.11037036776542664, "learning_rate": 6.70447792786083e-05, "loss": 0.0388, "step": 67170 }, { "epoch": 1.3424723232484712, "grad_norm": 0.14815907180309296, "learning_rate": 6.703506373398364e-05, "loss": 0.0368, "step": 67180 }, { "epoch": 1.3426721553894727, "grad_norm": 0.14357289671897888, "learning_rate": 6.702534746165235e-05, "loss": 0.0392, "step": 67190 }, { "epoch": 1.3428719875304744, "grad_norm": 0.12644341588020325, "learning_rate": 6.70156304620295e-05, "loss": 0.0285, "step": 67200 }, { "epoch": 1.343071819671476, "grad_norm": 0.12931875884532928, "learning_rate": 6.700591273553017e-05, "loss": 0.0343, "step": 67210 }, { "epoch": 1.3432716518124774, "grad_norm": 0.17372436821460724, "learning_rate": 6.69961942825695e-05, "loss": 0.0279, "step": 67220 }, { "epoch": 1.3434714839534792, "grad_norm": 0.13347505033016205, "learning_rate": 6.698647510356262e-05, "loss": 0.0318, "step": 67230 }, { "epoch": 1.3436713160944806, "grad_norm": 0.12305126339197159, "learning_rate": 6.697675519892475e-05, "loss": 0.0382, "step": 67240 }, { "epoch": 1.3438711482354821, "grad_norm": 0.13445357978343964, "learning_rate": 6.696703456907105e-05, "loss": 0.0319, "step": 67250 }, { "epoch": 1.3440709803764839, "grad_norm": 0.12253957986831665, "learning_rate": 6.695731321441682e-05, "loss": 0.0286, "step": 67260 }, { "epoch": 1.3442708125174854, "grad_norm": 0.1689126342535019, "learning_rate": 6.694759113537732e-05, "loss": 0.0347, "step": 67270 }, { "epoch": 1.3444706446584869, "grad_norm": 0.13944663107395172, "learning_rate": 6.693786833236787e-05, "loss": 0.0312, "step": 67280 }, { "epoch": 1.3446704767994884, "grad_norm": 0.126290425658226, "learning_rate": 6.692814480580379e-05, "loss": 0.0329, "step": 67290 }, { "epoch": 1.3448703089404899, "grad_norm": 0.1481887549161911, "learning_rate": 6.691842055610048e-05, "loss": 0.0413, "step": 67300 }, { "epoch": 1.3450701410814916, "grad_norm": 0.16177591681480408, "learning_rate": 6.69086955836733e-05, "loss": 0.0318, "step": 67310 }, { "epoch": 1.345269973222493, "grad_norm": 0.12021104991436005, "learning_rate": 6.689896988893774e-05, "loss": 0.037, "step": 67320 }, { "epoch": 1.3454698053634946, "grad_norm": 0.08687978237867355, "learning_rate": 6.68892434723092e-05, "loss": 0.031, "step": 67330 }, { "epoch": 1.3456696375044963, "grad_norm": 0.09797686338424683, "learning_rate": 6.687951633420321e-05, "loss": 0.0324, "step": 67340 }, { "epoch": 1.3458694696454978, "grad_norm": 0.1280232071876526, "learning_rate": 6.68697884750353e-05, "loss": 0.0304, "step": 67350 }, { "epoch": 1.3460693017864993, "grad_norm": 0.1401008516550064, "learning_rate": 6.686005989522099e-05, "loss": 0.0277, "step": 67360 }, { "epoch": 1.346269133927501, "grad_norm": 0.14872370660305023, "learning_rate": 6.685033059517593e-05, "loss": 0.0289, "step": 67370 }, { "epoch": 1.3464689660685025, "grad_norm": 0.14405840635299683, "learning_rate": 6.684060057531569e-05, "loss": 0.0377, "step": 67380 }, { "epoch": 1.346668798209504, "grad_norm": 0.12976215779781342, "learning_rate": 6.683086983605593e-05, "loss": 0.0314, "step": 67390 }, { "epoch": 1.3468686303505055, "grad_norm": 0.12591345608234406, "learning_rate": 6.682113837781234e-05, "loss": 0.0352, "step": 67400 }, { "epoch": 1.347068462491507, "grad_norm": 0.13204702734947205, "learning_rate": 6.681140620100059e-05, "loss": 0.0331, "step": 67410 }, { "epoch": 1.3472682946325087, "grad_norm": 0.1833423525094986, "learning_rate": 6.680167330603648e-05, "loss": 0.0314, "step": 67420 }, { "epoch": 1.3474681267735102, "grad_norm": 0.11304297298192978, "learning_rate": 6.679193969333575e-05, "loss": 0.0277, "step": 67430 }, { "epoch": 1.3476679589145117, "grad_norm": 0.12825307250022888, "learning_rate": 6.678220536331419e-05, "loss": 0.0338, "step": 67440 }, { "epoch": 1.3478677910555135, "grad_norm": 0.12709398567676544, "learning_rate": 6.677247031638765e-05, "loss": 0.0315, "step": 67450 }, { "epoch": 1.348067623196515, "grad_norm": 0.13176876306533813, "learning_rate": 6.676273455297198e-05, "loss": 0.024, "step": 67460 }, { "epoch": 1.3482674553375165, "grad_norm": 0.0897585079073906, "learning_rate": 6.67529980734831e-05, "loss": 0.0366, "step": 67470 }, { "epoch": 1.348467287478518, "grad_norm": 0.11169061064720154, "learning_rate": 6.674326087833691e-05, "loss": 0.0344, "step": 67480 }, { "epoch": 1.3486671196195197, "grad_norm": 0.12117468565702438, "learning_rate": 6.67335229679494e-05, "loss": 0.0271, "step": 67490 }, { "epoch": 1.3488669517605212, "grad_norm": 0.13535666465759277, "learning_rate": 6.672378434273649e-05, "loss": 0.0297, "step": 67500 }, { "epoch": 1.3490667839015227, "grad_norm": 0.13190792500972748, "learning_rate": 6.671404500311426e-05, "loss": 0.0371, "step": 67510 }, { "epoch": 1.3492666160425242, "grad_norm": 0.1353430300951004, "learning_rate": 6.670430494949872e-05, "loss": 0.0347, "step": 67520 }, { "epoch": 1.349466448183526, "grad_norm": 0.11008729040622711, "learning_rate": 6.669456418230596e-05, "loss": 0.0322, "step": 67530 }, { "epoch": 1.3496662803245274, "grad_norm": 0.14444252848625183, "learning_rate": 6.66848227019521e-05, "loss": 0.0318, "step": 67540 }, { "epoch": 1.349866112465529, "grad_norm": 0.15593977272510529, "learning_rate": 6.667508050885325e-05, "loss": 0.0344, "step": 67550 }, { "epoch": 1.3500659446065306, "grad_norm": 0.11264276504516602, "learning_rate": 6.66653376034256e-05, "loss": 0.0279, "step": 67560 }, { "epoch": 1.3502657767475321, "grad_norm": 0.11026224493980408, "learning_rate": 6.665559398608535e-05, "loss": 0.034, "step": 67570 }, { "epoch": 1.3504656088885336, "grad_norm": 0.12820494174957275, "learning_rate": 6.664584965724872e-05, "loss": 0.0223, "step": 67580 }, { "epoch": 1.3506654410295351, "grad_norm": 0.12355169653892517, "learning_rate": 6.663610461733196e-05, "loss": 0.0365, "step": 67590 }, { "epoch": 1.3508652731705366, "grad_norm": 0.14205241203308105, "learning_rate": 6.662635886675139e-05, "loss": 0.041, "step": 67600 }, { "epoch": 1.3510651053115383, "grad_norm": 0.14288754761219025, "learning_rate": 6.66166124059233e-05, "loss": 0.0304, "step": 67610 }, { "epoch": 1.3512649374525398, "grad_norm": 0.10198413580656052, "learning_rate": 6.660686523526406e-05, "loss": 0.0295, "step": 67620 }, { "epoch": 1.3514647695935413, "grad_norm": 0.12115375697612762, "learning_rate": 6.659711735519003e-05, "loss": 0.0356, "step": 67630 }, { "epoch": 1.351664601734543, "grad_norm": 0.0943731889128685, "learning_rate": 6.658736876611765e-05, "loss": 0.0266, "step": 67640 }, { "epoch": 1.3518644338755446, "grad_norm": 0.1442955732345581, "learning_rate": 6.657761946846335e-05, "loss": 0.026, "step": 67650 }, { "epoch": 1.352064266016546, "grad_norm": 0.13542892038822174, "learning_rate": 6.65678694626436e-05, "loss": 0.0308, "step": 67660 }, { "epoch": 1.3522640981575478, "grad_norm": 0.13276070356369019, "learning_rate": 6.65581187490749e-05, "loss": 0.0379, "step": 67670 }, { "epoch": 1.3524639302985493, "grad_norm": 0.10495070368051529, "learning_rate": 6.654836732817378e-05, "loss": 0.0321, "step": 67680 }, { "epoch": 1.3526637624395508, "grad_norm": 0.11061441153287888, "learning_rate": 6.653861520035681e-05, "loss": 0.0298, "step": 67690 }, { "epoch": 1.3528635945805523, "grad_norm": 0.13956952095031738, "learning_rate": 6.65288623660406e-05, "loss": 0.0318, "step": 67700 }, { "epoch": 1.3530634267215538, "grad_norm": 0.14601175487041473, "learning_rate": 6.651910882564171e-05, "loss": 0.0308, "step": 67710 }, { "epoch": 1.3532632588625555, "grad_norm": 0.16927596926689148, "learning_rate": 6.650935457957688e-05, "loss": 0.0268, "step": 67720 }, { "epoch": 1.353463091003557, "grad_norm": 0.11660136282444, "learning_rate": 6.649959962826273e-05, "loss": 0.0277, "step": 67730 }, { "epoch": 1.3536629231445585, "grad_norm": 0.12533415853977203, "learning_rate": 6.6489843972116e-05, "loss": 0.0272, "step": 67740 }, { "epoch": 1.3538627552855602, "grad_norm": 0.10598456859588623, "learning_rate": 6.648008761155344e-05, "loss": 0.0289, "step": 67750 }, { "epoch": 1.3540625874265617, "grad_norm": 0.12988823652267456, "learning_rate": 6.647033054699179e-05, "loss": 0.0337, "step": 67760 }, { "epoch": 1.3542624195675632, "grad_norm": 0.10954828560352325, "learning_rate": 6.64605727788479e-05, "loss": 0.0289, "step": 67770 }, { "epoch": 1.354462251708565, "grad_norm": 0.11541111022233963, "learning_rate": 6.645081430753856e-05, "loss": 0.029, "step": 67780 }, { "epoch": 1.3546620838495664, "grad_norm": 0.22752371430397034, "learning_rate": 6.644105513348066e-05, "loss": 0.0322, "step": 67790 }, { "epoch": 1.354861915990568, "grad_norm": 0.16713784635066986, "learning_rate": 6.64312952570911e-05, "loss": 0.0348, "step": 67800 }, { "epoch": 1.3550617481315694, "grad_norm": 0.13776405155658722, "learning_rate": 6.642153467878677e-05, "loss": 0.0261, "step": 67810 }, { "epoch": 1.355261580272571, "grad_norm": 0.15068216621875763, "learning_rate": 6.641177339898465e-05, "loss": 0.0352, "step": 67820 }, { "epoch": 1.3554614124135727, "grad_norm": 0.12942646443843842, "learning_rate": 6.640201141810174e-05, "loss": 0.0289, "step": 67830 }, { "epoch": 1.3556612445545742, "grad_norm": 0.13835109770298004, "learning_rate": 6.639224873655502e-05, "loss": 0.033, "step": 67840 }, { "epoch": 1.3558610766955757, "grad_norm": 0.13485494256019592, "learning_rate": 6.638248535476155e-05, "loss": 0.0282, "step": 67850 }, { "epoch": 1.3560609088365774, "grad_norm": 0.15088176727294922, "learning_rate": 6.637272127313838e-05, "loss": 0.0267, "step": 67860 }, { "epoch": 1.3562607409775789, "grad_norm": 0.14038459956645966, "learning_rate": 6.636295649210267e-05, "loss": 0.0271, "step": 67870 }, { "epoch": 1.3564605731185804, "grad_norm": 0.1385839879512787, "learning_rate": 6.63531910120715e-05, "loss": 0.031, "step": 67880 }, { "epoch": 1.3566604052595819, "grad_norm": 0.0910223200917244, "learning_rate": 6.634342483346205e-05, "loss": 0.0366, "step": 67890 }, { "epoch": 1.3568602374005836, "grad_norm": 0.10884310305118561, "learning_rate": 6.633365795669152e-05, "loss": 0.0341, "step": 67900 }, { "epoch": 1.357060069541585, "grad_norm": 0.10801970213651657, "learning_rate": 6.632389038217712e-05, "loss": 0.0251, "step": 67910 }, { "epoch": 1.3572599016825866, "grad_norm": 0.14595773816108704, "learning_rate": 6.631412211033611e-05, "loss": 0.0356, "step": 67920 }, { "epoch": 1.357459733823588, "grad_norm": 0.1325804591178894, "learning_rate": 6.630435314158579e-05, "loss": 0.0383, "step": 67930 }, { "epoch": 1.3576595659645898, "grad_norm": 0.14175945520401, "learning_rate": 6.629458347634345e-05, "loss": 0.0367, "step": 67940 }, { "epoch": 1.3578593981055913, "grad_norm": 0.1246587485074997, "learning_rate": 6.628481311502645e-05, "loss": 0.0272, "step": 67950 }, { "epoch": 1.3580592302465928, "grad_norm": 0.12581297755241394, "learning_rate": 6.627504205805213e-05, "loss": 0.038, "step": 67960 }, { "epoch": 1.3582590623875945, "grad_norm": 0.14200444519519806, "learning_rate": 6.626527030583791e-05, "loss": 0.0261, "step": 67970 }, { "epoch": 1.358458894528596, "grad_norm": 0.14222756028175354, "learning_rate": 6.625549785880123e-05, "loss": 0.0293, "step": 67980 }, { "epoch": 1.3586587266695975, "grad_norm": 0.11706015467643738, "learning_rate": 6.624572471735954e-05, "loss": 0.0388, "step": 67990 }, { "epoch": 1.358858558810599, "grad_norm": 0.1018613800406456, "learning_rate": 6.623595088193034e-05, "loss": 0.0399, "step": 68000 }, { "epoch": 1.3590583909516005, "grad_norm": 0.11692109704017639, "learning_rate": 6.622617635293113e-05, "loss": 0.0382, "step": 68010 }, { "epoch": 1.3592582230926022, "grad_norm": 0.10793741792440414, "learning_rate": 6.62164011307795e-05, "loss": 0.0289, "step": 68020 }, { "epoch": 1.3594580552336037, "grad_norm": 0.15464448928833008, "learning_rate": 6.6206625215893e-05, "loss": 0.0313, "step": 68030 }, { "epoch": 1.3596578873746052, "grad_norm": 0.1741349846124649, "learning_rate": 6.619684860868922e-05, "loss": 0.0383, "step": 68040 }, { "epoch": 1.359857719515607, "grad_norm": 0.15309183299541473, "learning_rate": 6.618707130958585e-05, "loss": 0.0405, "step": 68050 }, { "epoch": 1.3600575516566085, "grad_norm": 0.12197203934192657, "learning_rate": 6.61772933190005e-05, "loss": 0.0326, "step": 68060 }, { "epoch": 1.36025738379761, "grad_norm": 0.15920130908489227, "learning_rate": 6.616751463735093e-05, "loss": 0.0263, "step": 68070 }, { "epoch": 1.3604572159386117, "grad_norm": 0.13610674440860748, "learning_rate": 6.61577352650548e-05, "loss": 0.0306, "step": 68080 }, { "epoch": 1.3606570480796132, "grad_norm": 0.11495237052440643, "learning_rate": 6.614795520252994e-05, "loss": 0.0309, "step": 68090 }, { "epoch": 1.3608568802206147, "grad_norm": 0.14257220923900604, "learning_rate": 6.613817445019407e-05, "loss": 0.0272, "step": 68100 }, { "epoch": 1.3610567123616162, "grad_norm": 0.1171519011259079, "learning_rate": 6.612839300846505e-05, "loss": 0.0272, "step": 68110 }, { "epoch": 1.3612565445026177, "grad_norm": 0.09127172082662582, "learning_rate": 6.61186108777607e-05, "loss": 0.0342, "step": 68120 }, { "epoch": 1.3614563766436194, "grad_norm": 0.11262550204992294, "learning_rate": 6.610882805849891e-05, "loss": 0.0327, "step": 68130 }, { "epoch": 1.361656208784621, "grad_norm": 0.20143592357635498, "learning_rate": 6.609904455109758e-05, "loss": 0.035, "step": 68140 }, { "epoch": 1.3618560409256224, "grad_norm": 0.11440219730138779, "learning_rate": 6.608926035597464e-05, "loss": 0.0318, "step": 68150 }, { "epoch": 1.3620558730666241, "grad_norm": 0.13710767030715942, "learning_rate": 6.607947547354804e-05, "loss": 0.0257, "step": 68160 }, { "epoch": 1.3622557052076256, "grad_norm": 0.09584128111600876, "learning_rate": 6.606968990423581e-05, "loss": 0.0304, "step": 68170 }, { "epoch": 1.3624555373486271, "grad_norm": 0.1283383071422577, "learning_rate": 6.605990364845593e-05, "loss": 0.0311, "step": 68180 }, { "epoch": 1.3626553694896288, "grad_norm": 0.12314490228891373, "learning_rate": 6.605011670662649e-05, "loss": 0.0277, "step": 68190 }, { "epoch": 1.3628552016306303, "grad_norm": 0.1257096529006958, "learning_rate": 6.604032907916553e-05, "loss": 0.0299, "step": 68200 }, { "epoch": 1.3630550337716318, "grad_norm": 0.15399514138698578, "learning_rate": 6.603054076649117e-05, "loss": 0.0375, "step": 68210 }, { "epoch": 1.3632548659126333, "grad_norm": 0.1316831409931183, "learning_rate": 6.602075176902158e-05, "loss": 0.0347, "step": 68220 }, { "epoch": 1.3634546980536348, "grad_norm": 0.12903976440429688, "learning_rate": 6.60109620871749e-05, "loss": 0.0429, "step": 68230 }, { "epoch": 1.3636545301946366, "grad_norm": 0.11554673314094543, "learning_rate": 6.600117172136934e-05, "loss": 0.0259, "step": 68240 }, { "epoch": 1.363854362335638, "grad_norm": 0.12073888629674911, "learning_rate": 6.59913806720231e-05, "loss": 0.0266, "step": 68250 }, { "epoch": 1.3640541944766396, "grad_norm": 0.13111092150211334, "learning_rate": 6.598158893955446e-05, "loss": 0.0276, "step": 68260 }, { "epoch": 1.3642540266176413, "grad_norm": 0.12250477820634842, "learning_rate": 6.597179652438173e-05, "loss": 0.0257, "step": 68270 }, { "epoch": 1.3644538587586428, "grad_norm": 0.14037206768989563, "learning_rate": 6.596200342692315e-05, "loss": 0.0326, "step": 68280 }, { "epoch": 1.3646536908996443, "grad_norm": 0.12309978902339935, "learning_rate": 6.595220964759713e-05, "loss": 0.0354, "step": 68290 }, { "epoch": 1.3648535230406458, "grad_norm": 0.14269889891147614, "learning_rate": 6.5942415186822e-05, "loss": 0.0375, "step": 68300 }, { "epoch": 1.3650533551816475, "grad_norm": 0.10447581112384796, "learning_rate": 6.59326200450162e-05, "loss": 0.0289, "step": 68310 }, { "epoch": 1.365253187322649, "grad_norm": 0.12795689702033997, "learning_rate": 6.592282422259813e-05, "loss": 0.0333, "step": 68320 }, { "epoch": 1.3654530194636505, "grad_norm": 0.1275523602962494, "learning_rate": 6.591302771998626e-05, "loss": 0.036, "step": 68330 }, { "epoch": 1.365652851604652, "grad_norm": 0.1338297724723816, "learning_rate": 6.59032305375991e-05, "loss": 0.0307, "step": 68340 }, { "epoch": 1.3658526837456537, "grad_norm": 0.13483017683029175, "learning_rate": 6.589343267585513e-05, "loss": 0.0325, "step": 68350 }, { "epoch": 1.3660525158866552, "grad_norm": 0.11675120145082474, "learning_rate": 6.588363413517291e-05, "loss": 0.0313, "step": 68360 }, { "epoch": 1.3662523480276567, "grad_norm": 0.12641440331935883, "learning_rate": 6.587383491597102e-05, "loss": 0.0322, "step": 68370 }, { "epoch": 1.3664521801686584, "grad_norm": 0.15042392909526825, "learning_rate": 6.586403501866806e-05, "loss": 0.0298, "step": 68380 }, { "epoch": 1.36665201230966, "grad_norm": 0.11698594689369202, "learning_rate": 6.585423444368268e-05, "loss": 0.0299, "step": 68390 }, { "epoch": 1.3668518444506614, "grad_norm": 0.14112289249897003, "learning_rate": 6.584443319143353e-05, "loss": 0.0343, "step": 68400 }, { "epoch": 1.367051676591663, "grad_norm": 0.13568301498889923, "learning_rate": 6.583463126233928e-05, "loss": 0.0444, "step": 68410 }, { "epoch": 1.3672515087326644, "grad_norm": 0.12560218572616577, "learning_rate": 6.58248286568187e-05, "loss": 0.0355, "step": 68420 }, { "epoch": 1.3674513408736662, "grad_norm": 0.09787115454673767, "learning_rate": 6.58150253752905e-05, "loss": 0.0422, "step": 68430 }, { "epoch": 1.3676511730146677, "grad_norm": 0.10332468152046204, "learning_rate": 6.580522141817345e-05, "loss": 0.0277, "step": 68440 }, { "epoch": 1.3678510051556692, "grad_norm": 0.1091519370675087, "learning_rate": 6.57954167858864e-05, "loss": 0.0275, "step": 68450 }, { "epoch": 1.3680508372966709, "grad_norm": 0.11853727698326111, "learning_rate": 6.578561147884815e-05, "loss": 0.0259, "step": 68460 }, { "epoch": 1.3682506694376724, "grad_norm": 0.16083954274654388, "learning_rate": 6.57758054974776e-05, "loss": 0.0353, "step": 68470 }, { "epoch": 1.3684505015786739, "grad_norm": 0.11624085903167725, "learning_rate": 6.576599884219359e-05, "loss": 0.0257, "step": 68480 }, { "epoch": 1.3686503337196756, "grad_norm": 0.1500941812992096, "learning_rate": 6.575619151341508e-05, "loss": 0.0396, "step": 68490 }, { "epoch": 1.368850165860677, "grad_norm": 0.11625123769044876, "learning_rate": 6.574638351156101e-05, "loss": 0.0289, "step": 68500 }, { "epoch": 1.3690499980016786, "grad_norm": 0.13802701234817505, "learning_rate": 6.573657483705038e-05, "loss": 0.0319, "step": 68510 }, { "epoch": 1.36924983014268, "grad_norm": 0.12093557417392731, "learning_rate": 6.57267654903022e-05, "loss": 0.0304, "step": 68520 }, { "epoch": 1.3694496622836816, "grad_norm": 0.11143049597740173, "learning_rate": 6.571695547173543e-05, "loss": 0.0302, "step": 68530 }, { "epoch": 1.3696494944246833, "grad_norm": 0.12612757086753845, "learning_rate": 6.570714478176924e-05, "loss": 0.0402, "step": 68540 }, { "epoch": 1.3698493265656848, "grad_norm": 0.1379699409008026, "learning_rate": 6.569733342082267e-05, "loss": 0.0328, "step": 68550 }, { "epoch": 1.3700491587066863, "grad_norm": 0.09903238713741302, "learning_rate": 6.568752138931484e-05, "loss": 0.0295, "step": 68560 }, { "epoch": 1.370248990847688, "grad_norm": 0.16048185527324677, "learning_rate": 6.567770868766492e-05, "loss": 0.0387, "step": 68570 }, { "epoch": 1.3704488229886895, "grad_norm": 0.1491241306066513, "learning_rate": 6.56678953162921e-05, "loss": 0.033, "step": 68580 }, { "epoch": 1.370648655129691, "grad_norm": 0.159645214676857, "learning_rate": 6.565808127561555e-05, "loss": 0.0273, "step": 68590 }, { "epoch": 1.3708484872706928, "grad_norm": 0.12892906367778778, "learning_rate": 6.564826656605457e-05, "loss": 0.0298, "step": 68600 }, { "epoch": 1.3710483194116942, "grad_norm": 0.13725966215133667, "learning_rate": 6.563845118802836e-05, "loss": 0.0346, "step": 68610 }, { "epoch": 1.3712481515526957, "grad_norm": 0.15586824715137482, "learning_rate": 6.562863514195626e-05, "loss": 0.0293, "step": 68620 }, { "epoch": 1.3714479836936972, "grad_norm": 0.08907197415828705, "learning_rate": 6.561881842825757e-05, "loss": 0.0298, "step": 68630 }, { "epoch": 1.3716478158346987, "grad_norm": 0.13951613008975983, "learning_rate": 6.560900104735163e-05, "loss": 0.0281, "step": 68640 }, { "epoch": 1.3718476479757005, "grad_norm": 0.19365717470645905, "learning_rate": 6.559918299965787e-05, "loss": 0.0344, "step": 68650 }, { "epoch": 1.372047480116702, "grad_norm": 0.16547442972660065, "learning_rate": 6.558936428559565e-05, "loss": 0.0344, "step": 68660 }, { "epoch": 1.3722473122577035, "grad_norm": 0.14064010977745056, "learning_rate": 6.557954490558444e-05, "loss": 0.026, "step": 68670 }, { "epoch": 1.3724471443987052, "grad_norm": 0.14463011920452118, "learning_rate": 6.556972486004367e-05, "loss": 0.0351, "step": 68680 }, { "epoch": 1.3726469765397067, "grad_norm": 0.18042586743831635, "learning_rate": 6.555990414939289e-05, "loss": 0.0348, "step": 68690 }, { "epoch": 1.3728468086807082, "grad_norm": 0.17010411620140076, "learning_rate": 6.555008277405156e-05, "loss": 0.0359, "step": 68700 }, { "epoch": 1.3730466408217097, "grad_norm": 0.18211540579795837, "learning_rate": 6.554026073443929e-05, "loss": 0.0398, "step": 68710 }, { "epoch": 1.3732464729627112, "grad_norm": 0.17349417507648468, "learning_rate": 6.553043803097562e-05, "loss": 0.0324, "step": 68720 }, { "epoch": 1.373446305103713, "grad_norm": 0.1511690616607666, "learning_rate": 6.552061466408014e-05, "loss": 0.0296, "step": 68730 }, { "epoch": 1.3736461372447144, "grad_norm": 0.1307605504989624, "learning_rate": 6.551079063417254e-05, "loss": 0.0303, "step": 68740 }, { "epoch": 1.373845969385716, "grad_norm": 0.12675318121910095, "learning_rate": 6.550096594167245e-05, "loss": 0.0334, "step": 68750 }, { "epoch": 1.3740458015267176, "grad_norm": 0.13317368924617767, "learning_rate": 6.549114058699958e-05, "loss": 0.0359, "step": 68760 }, { "epoch": 1.3742456336677191, "grad_norm": 0.11158277094364166, "learning_rate": 6.548131457057363e-05, "loss": 0.0308, "step": 68770 }, { "epoch": 1.3744454658087206, "grad_norm": 0.11421818286180496, "learning_rate": 6.547148789281436e-05, "loss": 0.0289, "step": 68780 }, { "epoch": 1.3746452979497223, "grad_norm": 0.1424238681793213, "learning_rate": 6.546166055414155e-05, "loss": 0.0334, "step": 68790 }, { "epoch": 1.3748451300907238, "grad_norm": 0.1163037121295929, "learning_rate": 6.545183255497503e-05, "loss": 0.0378, "step": 68800 }, { "epoch": 1.3750449622317253, "grad_norm": 0.12797218561172485, "learning_rate": 6.544200389573458e-05, "loss": 0.0271, "step": 68810 }, { "epoch": 1.3752447943727268, "grad_norm": 0.14355415105819702, "learning_rate": 6.54321745768401e-05, "loss": 0.0324, "step": 68820 }, { "epoch": 1.3754446265137283, "grad_norm": 0.1387479305267334, "learning_rate": 6.542234459871149e-05, "loss": 0.0362, "step": 68830 }, { "epoch": 1.37564445865473, "grad_norm": 0.15382710099220276, "learning_rate": 6.541251396176861e-05, "loss": 0.0262, "step": 68840 }, { "epoch": 1.3758442907957316, "grad_norm": 0.13184398412704468, "learning_rate": 6.540268266643148e-05, "loss": 0.0404, "step": 68850 }, { "epoch": 1.376044122936733, "grad_norm": 0.14022156596183777, "learning_rate": 6.539285071312004e-05, "loss": 0.0327, "step": 68860 }, { "epoch": 1.3762439550777348, "grad_norm": 0.14761137962341309, "learning_rate": 6.53830181022543e-05, "loss": 0.0263, "step": 68870 }, { "epoch": 1.3764437872187363, "grad_norm": 0.15047144889831543, "learning_rate": 6.537318483425425e-05, "loss": 0.0437, "step": 68880 }, { "epoch": 1.3766436193597378, "grad_norm": 0.1263745129108429, "learning_rate": 6.536335090954002e-05, "loss": 0.0314, "step": 68890 }, { "epoch": 1.3768434515007395, "grad_norm": 0.15376512706279755, "learning_rate": 6.535351632853166e-05, "loss": 0.0358, "step": 68900 }, { "epoch": 1.377043283641741, "grad_norm": 0.13411785662174225, "learning_rate": 6.534368109164927e-05, "loss": 0.0401, "step": 68910 }, { "epoch": 1.3772431157827425, "grad_norm": 0.13407042622566223, "learning_rate": 6.533384519931302e-05, "loss": 0.0301, "step": 68920 }, { "epoch": 1.377442947923744, "grad_norm": 0.11210381239652634, "learning_rate": 6.532400865194307e-05, "loss": 0.0348, "step": 68930 }, { "epoch": 1.3776427800647455, "grad_norm": 0.1290895640850067, "learning_rate": 6.531417144995964e-05, "loss": 0.0288, "step": 68940 }, { "epoch": 1.3778426122057472, "grad_norm": 0.11426477134227753, "learning_rate": 6.530433359378292e-05, "loss": 0.0332, "step": 68950 }, { "epoch": 1.3780424443467487, "grad_norm": 0.11444157361984253, "learning_rate": 6.529449508383318e-05, "loss": 0.0306, "step": 68960 }, { "epoch": 1.3782422764877502, "grad_norm": 0.1269838809967041, "learning_rate": 6.52846559205307e-05, "loss": 0.0252, "step": 68970 }, { "epoch": 1.378442108628752, "grad_norm": 0.12513519823551178, "learning_rate": 6.527481610429582e-05, "loss": 0.0522, "step": 68980 }, { "epoch": 1.3786419407697534, "grad_norm": 0.09866508841514587, "learning_rate": 6.526497563554886e-05, "loss": 0.0294, "step": 68990 }, { "epoch": 1.378841772910755, "grad_norm": 0.1322220414876938, "learning_rate": 6.525513451471017e-05, "loss": 0.0282, "step": 69000 }, { "epoch": 1.3790416050517567, "grad_norm": 0.1407204270362854, "learning_rate": 6.524529274220017e-05, "loss": 0.0404, "step": 69010 }, { "epoch": 1.3792414371927582, "grad_norm": 0.1259070336818695, "learning_rate": 6.523545031843926e-05, "loss": 0.0247, "step": 69020 }, { "epoch": 1.3794412693337597, "grad_norm": 0.14723922312259674, "learning_rate": 6.52256072438479e-05, "loss": 0.0314, "step": 69030 }, { "epoch": 1.3796411014747612, "grad_norm": 0.13379733264446259, "learning_rate": 6.521576351884657e-05, "loss": 0.0339, "step": 69040 }, { "epoch": 1.3798409336157627, "grad_norm": 0.12456698715686798, "learning_rate": 6.520591914385576e-05, "loss": 0.0323, "step": 69050 }, { "epoch": 1.3800407657567644, "grad_norm": 0.16247384250164032, "learning_rate": 6.519607411929605e-05, "loss": 0.0276, "step": 69060 }, { "epoch": 1.3802405978977659, "grad_norm": 0.11643655598163605, "learning_rate": 6.518622844558795e-05, "loss": 0.0238, "step": 69070 }, { "epoch": 1.3804404300387674, "grad_norm": 0.14704309403896332, "learning_rate": 6.517638212315208e-05, "loss": 0.0282, "step": 69080 }, { "epoch": 1.380640262179769, "grad_norm": 0.09221573173999786, "learning_rate": 6.516653515240903e-05, "loss": 0.0297, "step": 69090 }, { "epoch": 1.3808400943207706, "grad_norm": 0.13619773089885712, "learning_rate": 6.515668753377947e-05, "loss": 0.0351, "step": 69100 }, { "epoch": 1.381039926461772, "grad_norm": 0.1249098926782608, "learning_rate": 6.514683926768405e-05, "loss": 0.038, "step": 69110 }, { "epoch": 1.3812397586027736, "grad_norm": 0.12452677637338638, "learning_rate": 6.51369903545435e-05, "loss": 0.0286, "step": 69120 }, { "epoch": 1.381439590743775, "grad_norm": 0.1305142492055893, "learning_rate": 6.512714079477852e-05, "loss": 0.0319, "step": 69130 }, { "epoch": 1.3816394228847768, "grad_norm": 0.1450275182723999, "learning_rate": 6.511729058880988e-05, "loss": 0.0257, "step": 69140 }, { "epoch": 1.3818392550257783, "grad_norm": 0.11560410261154175, "learning_rate": 6.510743973705836e-05, "loss": 0.0329, "step": 69150 }, { "epoch": 1.3820390871667798, "grad_norm": 0.12441606819629669, "learning_rate": 6.509758823994477e-05, "loss": 0.027, "step": 69160 }, { "epoch": 1.3822389193077815, "grad_norm": 0.16360227763652802, "learning_rate": 6.508773609788994e-05, "loss": 0.0307, "step": 69170 }, { "epoch": 1.382438751448783, "grad_norm": 0.15110795199871063, "learning_rate": 6.507788331131473e-05, "loss": 0.0425, "step": 69180 }, { "epoch": 1.3826385835897845, "grad_norm": 0.13032284379005432, "learning_rate": 6.506802988064004e-05, "loss": 0.0362, "step": 69190 }, { "epoch": 1.3828384157307863, "grad_norm": 0.1375359147787094, "learning_rate": 6.505817580628682e-05, "loss": 0.0257, "step": 69200 }, { "epoch": 1.3830382478717878, "grad_norm": 0.15822045505046844, "learning_rate": 6.504832108867598e-05, "loss": 0.0333, "step": 69210 }, { "epoch": 1.3832380800127893, "grad_norm": 0.12144812941551208, "learning_rate": 6.503846572822851e-05, "loss": 0.0294, "step": 69220 }, { "epoch": 1.3834379121537907, "grad_norm": 0.14994941651821136, "learning_rate": 6.50286097253654e-05, "loss": 0.0372, "step": 69230 }, { "epoch": 1.3836377442947922, "grad_norm": 0.1505212038755417, "learning_rate": 6.501875308050768e-05, "loss": 0.0331, "step": 69240 }, { "epoch": 1.383837576435794, "grad_norm": 0.0996822938323021, "learning_rate": 6.500889579407645e-05, "loss": 0.0422, "step": 69250 }, { "epoch": 1.3840374085767955, "grad_norm": 0.16671381890773773, "learning_rate": 6.499903786649276e-05, "loss": 0.0305, "step": 69260 }, { "epoch": 1.384237240717797, "grad_norm": 0.1286957561969757, "learning_rate": 6.498917929817772e-05, "loss": 0.0309, "step": 69270 }, { "epoch": 1.3844370728587987, "grad_norm": 0.12195561081171036, "learning_rate": 6.497932008955247e-05, "loss": 0.0301, "step": 69280 }, { "epoch": 1.3846369049998002, "grad_norm": 0.1394473910331726, "learning_rate": 6.496946024103819e-05, "loss": 0.0326, "step": 69290 }, { "epoch": 1.3848367371408017, "grad_norm": 0.1940182000398636, "learning_rate": 6.495959975305606e-05, "loss": 0.0329, "step": 69300 }, { "epoch": 1.3850365692818034, "grad_norm": 0.16207647323608398, "learning_rate": 6.494973862602732e-05, "loss": 0.0291, "step": 69310 }, { "epoch": 1.385236401422805, "grad_norm": 0.1299373060464859, "learning_rate": 6.493987686037321e-05, "loss": 0.0343, "step": 69320 }, { "epoch": 1.3854362335638064, "grad_norm": 0.1927853375673294, "learning_rate": 6.493001445651498e-05, "loss": 0.0288, "step": 69330 }, { "epoch": 1.385636065704808, "grad_norm": 0.16413359344005585, "learning_rate": 6.4920151414874e-05, "loss": 0.0247, "step": 69340 }, { "epoch": 1.3858358978458094, "grad_norm": 0.10809504240751266, "learning_rate": 6.491028773587153e-05, "loss": 0.0292, "step": 69350 }, { "epoch": 1.3860357299868111, "grad_norm": 0.13207757472991943, "learning_rate": 6.490042341992898e-05, "loss": 0.0299, "step": 69360 }, { "epoch": 1.3862355621278126, "grad_norm": 0.13632845878601074, "learning_rate": 6.489055846746772e-05, "loss": 0.0353, "step": 69370 }, { "epoch": 1.3864353942688141, "grad_norm": 0.16430440545082092, "learning_rate": 6.488069287890912e-05, "loss": 0.0309, "step": 69380 }, { "epoch": 1.3866352264098158, "grad_norm": 0.201628178358078, "learning_rate": 6.487082665467469e-05, "loss": 0.0259, "step": 69390 }, { "epoch": 1.3868350585508173, "grad_norm": 0.13699109852313995, "learning_rate": 6.486095979518585e-05, "loss": 0.0339, "step": 69400 }, { "epoch": 1.3870348906918188, "grad_norm": 0.12055742740631104, "learning_rate": 6.485109230086412e-05, "loss": 0.03, "step": 69410 }, { "epoch": 1.3872347228328206, "grad_norm": 0.168496236205101, "learning_rate": 6.484122417213098e-05, "loss": 0.0287, "step": 69420 }, { "epoch": 1.387434554973822, "grad_norm": 0.13463932275772095, "learning_rate": 6.483135540940803e-05, "loss": 0.0257, "step": 69430 }, { "epoch": 1.3876343871148236, "grad_norm": 0.1346854567527771, "learning_rate": 6.482148601311683e-05, "loss": 0.0329, "step": 69440 }, { "epoch": 1.387834219255825, "grad_norm": 0.14232675731182098, "learning_rate": 6.481161598367897e-05, "loss": 0.0349, "step": 69450 }, { "epoch": 1.3880340513968266, "grad_norm": 0.12442702800035477, "learning_rate": 6.480174532151608e-05, "loss": 0.0325, "step": 69460 }, { "epoch": 1.3882338835378283, "grad_norm": 0.11972256004810333, "learning_rate": 6.479187402704983e-05, "loss": 0.0379, "step": 69470 }, { "epoch": 1.3884337156788298, "grad_norm": 0.1308608204126358, "learning_rate": 6.478200210070187e-05, "loss": 0.0302, "step": 69480 }, { "epoch": 1.3886335478198313, "grad_norm": 0.0961887389421463, "learning_rate": 6.477212954289396e-05, "loss": 0.0331, "step": 69490 }, { "epoch": 1.388833379960833, "grad_norm": 0.12625358998775482, "learning_rate": 6.47622563540478e-05, "loss": 0.0256, "step": 69500 }, { "epoch": 1.3890332121018345, "grad_norm": 0.118476003408432, "learning_rate": 6.475238253458518e-05, "loss": 0.0257, "step": 69510 }, { "epoch": 1.389233044242836, "grad_norm": 0.1559763252735138, "learning_rate": 6.474250808492787e-05, "loss": 0.0389, "step": 69520 }, { "epoch": 1.3894328763838375, "grad_norm": 0.31880122423171997, "learning_rate": 6.47326330054977e-05, "loss": 0.0275, "step": 69530 }, { "epoch": 1.389632708524839, "grad_norm": 0.11161038279533386, "learning_rate": 6.472275729671653e-05, "loss": 0.0325, "step": 69540 }, { "epoch": 1.3898325406658407, "grad_norm": 0.12770220637321472, "learning_rate": 6.47128809590062e-05, "loss": 0.0334, "step": 69550 }, { "epoch": 1.3900323728068422, "grad_norm": 0.13765384256839752, "learning_rate": 6.470300399278861e-05, "loss": 0.035, "step": 69560 }, { "epoch": 1.3902322049478437, "grad_norm": 0.1822502166032791, "learning_rate": 6.469312639848573e-05, "loss": 0.0412, "step": 69570 }, { "epoch": 1.3904320370888454, "grad_norm": 0.13568758964538574, "learning_rate": 6.468324817651945e-05, "loss": 0.0321, "step": 69580 }, { "epoch": 1.390631869229847, "grad_norm": 0.11142931133508682, "learning_rate": 6.467336932731181e-05, "loss": 0.0361, "step": 69590 }, { "epoch": 1.3908317013708484, "grad_norm": 0.13550622761249542, "learning_rate": 6.466348985128477e-05, "loss": 0.0384, "step": 69600 }, { "epoch": 1.3910315335118502, "grad_norm": 0.12854650616645813, "learning_rate": 6.465360974886038e-05, "loss": 0.0276, "step": 69610 }, { "epoch": 1.3912313656528517, "grad_norm": 0.13499145209789276, "learning_rate": 6.464372902046071e-05, "loss": 0.0314, "step": 69620 }, { "epoch": 1.3914311977938532, "grad_norm": 0.16157560050487518, "learning_rate": 6.463384766650783e-05, "loss": 0.032, "step": 69630 }, { "epoch": 1.3916310299348547, "grad_norm": 0.14693759381771088, "learning_rate": 6.462396568742389e-05, "loss": 0.0348, "step": 69640 }, { "epoch": 1.3918308620758562, "grad_norm": 0.1596151888370514, "learning_rate": 6.461408308363095e-05, "loss": 0.0255, "step": 69650 }, { "epoch": 1.3920306942168579, "grad_norm": 0.12049528956413269, "learning_rate": 6.460419985555126e-05, "loss": 0.0262, "step": 69660 }, { "epoch": 1.3922305263578594, "grad_norm": 0.12313620001077652, "learning_rate": 6.459431600360697e-05, "loss": 0.0373, "step": 69670 }, { "epoch": 1.3924303584988609, "grad_norm": 0.11730866879224777, "learning_rate": 6.458443152822032e-05, "loss": 0.0354, "step": 69680 }, { "epoch": 1.3926301906398626, "grad_norm": 0.1320699155330658, "learning_rate": 6.457454642981354e-05, "loss": 0.0317, "step": 69690 }, { "epoch": 1.392830022780864, "grad_norm": 0.12188360840082169, "learning_rate": 6.456466070880891e-05, "loss": 0.0282, "step": 69700 }, { "epoch": 1.3930298549218656, "grad_norm": 0.1406412422657013, "learning_rate": 6.455477436562874e-05, "loss": 0.0332, "step": 69710 }, { "epoch": 1.3932296870628673, "grad_norm": 0.12033253908157349, "learning_rate": 6.454488740069535e-05, "loss": 0.0264, "step": 69720 }, { "epoch": 1.3934295192038688, "grad_norm": 0.1192159354686737, "learning_rate": 6.453499981443109e-05, "loss": 0.0254, "step": 69730 }, { "epoch": 1.3936293513448703, "grad_norm": 0.13439755141735077, "learning_rate": 6.452511160725833e-05, "loss": 0.0254, "step": 69740 }, { "epoch": 1.3938291834858718, "grad_norm": 0.14873889088630676, "learning_rate": 6.451522277959948e-05, "loss": 0.0283, "step": 69750 }, { "epoch": 1.3940290156268733, "grad_norm": 0.12922228872776031, "learning_rate": 6.450533333187699e-05, "loss": 0.0288, "step": 69760 }, { "epoch": 1.394228847767875, "grad_norm": 0.128789484500885, "learning_rate": 6.449544326451331e-05, "loss": 0.0325, "step": 69770 }, { "epoch": 1.3944286799088765, "grad_norm": 0.13236527144908905, "learning_rate": 6.44855525779309e-05, "loss": 0.0266, "step": 69780 }, { "epoch": 1.394628512049878, "grad_norm": 0.12775535881519318, "learning_rate": 6.447566127255231e-05, "loss": 0.0279, "step": 69790 }, { "epoch": 1.3948283441908798, "grad_norm": 0.13576386868953705, "learning_rate": 6.446576934880007e-05, "loss": 0.0272, "step": 69800 }, { "epoch": 1.3950281763318813, "grad_norm": 0.13653534650802612, "learning_rate": 6.445587680709673e-05, "loss": 0.0357, "step": 69810 }, { "epoch": 1.3952280084728828, "grad_norm": 0.13142448663711548, "learning_rate": 6.44459836478649e-05, "loss": 0.029, "step": 69820 }, { "epoch": 1.3954278406138845, "grad_norm": 0.11857228726148605, "learning_rate": 6.443608987152716e-05, "loss": 0.0344, "step": 69830 }, { "epoch": 1.395627672754886, "grad_norm": 0.13895083963871002, "learning_rate": 6.442619547850621e-05, "loss": 0.0435, "step": 69840 }, { "epoch": 1.3958275048958875, "grad_norm": 0.1148693859577179, "learning_rate": 6.441630046922467e-05, "loss": 0.0315, "step": 69850 }, { "epoch": 1.396027337036889, "grad_norm": 0.13555476069450378, "learning_rate": 6.440640484410527e-05, "loss": 0.0378, "step": 69860 }, { "epoch": 1.3962271691778905, "grad_norm": 0.13563062250614166, "learning_rate": 6.439650860357072e-05, "loss": 0.0301, "step": 69870 }, { "epoch": 1.3964270013188922, "grad_norm": 0.11136174947023392, "learning_rate": 6.438661174804377e-05, "loss": 0.0295, "step": 69880 }, { "epoch": 1.3966268334598937, "grad_norm": 0.11765558272600174, "learning_rate": 6.437671427794718e-05, "loss": 0.0329, "step": 69890 }, { "epoch": 1.3968266656008952, "grad_norm": 0.13440102338790894, "learning_rate": 6.436681619370377e-05, "loss": 0.0294, "step": 69900 }, { "epoch": 1.397026497741897, "grad_norm": 0.16767145693302155, "learning_rate": 6.435691749573639e-05, "loss": 0.0328, "step": 69910 }, { "epoch": 1.3972263298828984, "grad_norm": 0.13259559869766235, "learning_rate": 6.434701818446785e-05, "loss": 0.0314, "step": 69920 }, { "epoch": 1.3974261620239, "grad_norm": 0.17231443524360657, "learning_rate": 6.433711826032106e-05, "loss": 0.0387, "step": 69930 }, { "epoch": 1.3976259941649014, "grad_norm": 0.15175898373126984, "learning_rate": 6.43272177237189e-05, "loss": 0.0291, "step": 69940 }, { "epoch": 1.397825826305903, "grad_norm": 0.11895115673542023, "learning_rate": 6.431731657508433e-05, "loss": 0.0381, "step": 69950 }, { "epoch": 1.3980256584469046, "grad_norm": 0.11232726275920868, "learning_rate": 6.430741481484028e-05, "loss": 0.0276, "step": 69960 }, { "epoch": 1.3982254905879061, "grad_norm": 0.11587022989988327, "learning_rate": 6.429751244340978e-05, "loss": 0.0263, "step": 69970 }, { "epoch": 1.3984253227289076, "grad_norm": 0.12168988585472107, "learning_rate": 6.428760946121579e-05, "loss": 0.0241, "step": 69980 }, { "epoch": 1.3986251548699093, "grad_norm": 0.12040803581476212, "learning_rate": 6.42777058686814e-05, "loss": 0.0363, "step": 69990 }, { "epoch": 1.3988249870109108, "grad_norm": 0.11969791352748871, "learning_rate": 6.426780166622962e-05, "loss": 0.0284, "step": 70000 }, { "epoch": 1.3990248191519123, "grad_norm": 0.12795931100845337, "learning_rate": 6.425789685428357e-05, "loss": 0.0253, "step": 70010 }, { "epoch": 1.399224651292914, "grad_norm": 0.13200289011001587, "learning_rate": 6.424799143326636e-05, "loss": 0.0315, "step": 70020 }, { "epoch": 1.3994244834339156, "grad_norm": 0.11086355149745941, "learning_rate": 6.423808540360112e-05, "loss": 0.0289, "step": 70030 }, { "epoch": 1.399624315574917, "grad_norm": 0.23386962711811066, "learning_rate": 6.422817876571104e-05, "loss": 0.037, "step": 70040 }, { "epoch": 1.3998241477159186, "grad_norm": 0.12975235283374786, "learning_rate": 6.421827152001929e-05, "loss": 0.0353, "step": 70050 }, { "epoch": 1.40002397985692, "grad_norm": 0.14586198329925537, "learning_rate": 6.420836366694911e-05, "loss": 0.0324, "step": 70060 }, { "epoch": 1.4002238119979218, "grad_norm": 0.18565134704113007, "learning_rate": 6.419845520692373e-05, "loss": 0.0352, "step": 70070 }, { "epoch": 1.4004236441389233, "grad_norm": 0.1267816573381424, "learning_rate": 6.418854614036644e-05, "loss": 0.0329, "step": 70080 }, { "epoch": 1.4006234762799248, "grad_norm": 0.15841813385486603, "learning_rate": 6.417863646770052e-05, "loss": 0.0401, "step": 70090 }, { "epoch": 1.4008233084209265, "grad_norm": 0.21351785957813263, "learning_rate": 6.416872618934928e-05, "loss": 0.0364, "step": 70100 }, { "epoch": 1.401023140561928, "grad_norm": 0.2124907523393631, "learning_rate": 6.415881530573611e-05, "loss": 0.028, "step": 70110 }, { "epoch": 1.4012229727029295, "grad_norm": 0.15530133247375488, "learning_rate": 6.414890381728434e-05, "loss": 0.0363, "step": 70120 }, { "epoch": 1.4014228048439312, "grad_norm": 0.14948955178260803, "learning_rate": 6.41389917244174e-05, "loss": 0.0317, "step": 70130 }, { "epoch": 1.4016226369849327, "grad_norm": 0.14234046638011932, "learning_rate": 6.412907902755872e-05, "loss": 0.0368, "step": 70140 }, { "epoch": 1.4018224691259342, "grad_norm": 0.1274229884147644, "learning_rate": 6.411916572713173e-05, "loss": 0.031, "step": 70150 }, { "epoch": 1.4020223012669357, "grad_norm": 0.13043327629566193, "learning_rate": 6.410925182355991e-05, "loss": 0.0363, "step": 70160 }, { "epoch": 1.4022221334079372, "grad_norm": 0.14405660331249237, "learning_rate": 6.409933731726678e-05, "loss": 0.0312, "step": 70170 }, { "epoch": 1.402421965548939, "grad_norm": 0.12694016098976135, "learning_rate": 6.408942220867585e-05, "loss": 0.032, "step": 70180 }, { "epoch": 1.4026217976899404, "grad_norm": 0.09531378000974655, "learning_rate": 6.40795064982107e-05, "loss": 0.0316, "step": 70190 }, { "epoch": 1.402821629830942, "grad_norm": 0.15884657204151154, "learning_rate": 6.40695901862949e-05, "loss": 0.0339, "step": 70200 }, { "epoch": 1.4030214619719437, "grad_norm": 0.14141565561294556, "learning_rate": 6.405967327335205e-05, "loss": 0.0301, "step": 70210 }, { "epoch": 1.4032212941129452, "grad_norm": 0.14598828554153442, "learning_rate": 6.404975575980578e-05, "loss": 0.0374, "step": 70220 }, { "epoch": 1.4034211262539467, "grad_norm": 0.09923030436038971, "learning_rate": 6.403983764607974e-05, "loss": 0.03, "step": 70230 }, { "epoch": 1.4036209583949484, "grad_norm": 0.09680058062076569, "learning_rate": 6.402991893259765e-05, "loss": 0.0331, "step": 70240 }, { "epoch": 1.4038207905359499, "grad_norm": 0.12644405663013458, "learning_rate": 6.401999961978317e-05, "loss": 0.0381, "step": 70250 }, { "epoch": 1.4040206226769514, "grad_norm": 0.17286501824855804, "learning_rate": 6.40100797080601e-05, "loss": 0.0323, "step": 70260 }, { "epoch": 1.4042204548179529, "grad_norm": 0.14701248705387115, "learning_rate": 6.400015919785215e-05, "loss": 0.0368, "step": 70270 }, { "epoch": 1.4044202869589544, "grad_norm": 0.13347330689430237, "learning_rate": 6.399023808958311e-05, "loss": 0.0292, "step": 70280 }, { "epoch": 1.404620119099956, "grad_norm": 0.1670561283826828, "learning_rate": 6.39803163836768e-05, "loss": 0.0409, "step": 70290 }, { "epoch": 1.4048199512409576, "grad_norm": 0.11708080768585205, "learning_rate": 6.397039408055706e-05, "loss": 0.0316, "step": 70300 }, { "epoch": 1.405019783381959, "grad_norm": 0.10705360770225525, "learning_rate": 6.396047118064777e-05, "loss": 0.027, "step": 70310 }, { "epoch": 1.4052196155229608, "grad_norm": 0.1304769217967987, "learning_rate": 6.395054768437276e-05, "loss": 0.0343, "step": 70320 }, { "epoch": 1.4054194476639623, "grad_norm": 0.1351383924484253, "learning_rate": 6.3940623592156e-05, "loss": 0.0422, "step": 70330 }, { "epoch": 1.4056192798049638, "grad_norm": 0.14910995960235596, "learning_rate": 6.39306989044214e-05, "loss": 0.0348, "step": 70340 }, { "epoch": 1.4058191119459653, "grad_norm": 0.11148572713136673, "learning_rate": 6.392077362159295e-05, "loss": 0.0239, "step": 70350 }, { "epoch": 1.4060189440869668, "grad_norm": 0.1669536978006363, "learning_rate": 6.39108477440946e-05, "loss": 0.0321, "step": 70360 }, { "epoch": 1.4062187762279685, "grad_norm": 0.13177262246608734, "learning_rate": 6.390092127235042e-05, "loss": 0.0301, "step": 70370 }, { "epoch": 1.40641860836897, "grad_norm": 0.1405751258134842, "learning_rate": 6.38909942067844e-05, "loss": 0.0338, "step": 70380 }, { "epoch": 1.4066184405099715, "grad_norm": 0.16666924953460693, "learning_rate": 6.388106654782063e-05, "loss": 0.0316, "step": 70390 }, { "epoch": 1.4068182726509733, "grad_norm": 0.1068025454878807, "learning_rate": 6.387113829588321e-05, "loss": 0.0395, "step": 70400 }, { "epoch": 1.4070181047919748, "grad_norm": 0.14639757573604584, "learning_rate": 6.386120945139623e-05, "loss": 0.0356, "step": 70410 }, { "epoch": 1.4072179369329763, "grad_norm": 0.1090712696313858, "learning_rate": 6.385128001478386e-05, "loss": 0.0319, "step": 70420 }, { "epoch": 1.407417769073978, "grad_norm": 0.15051324665546417, "learning_rate": 6.384134998647023e-05, "loss": 0.032, "step": 70430 }, { "epoch": 1.4076176012149795, "grad_norm": 0.11743421107530594, "learning_rate": 6.383141936687957e-05, "loss": 0.028, "step": 70440 }, { "epoch": 1.407817433355981, "grad_norm": 0.08897141367197037, "learning_rate": 6.382148815643607e-05, "loss": 0.0235, "step": 70450 }, { "epoch": 1.4080172654969825, "grad_norm": 0.1230073794722557, "learning_rate": 6.381155635556401e-05, "loss": 0.0275, "step": 70460 }, { "epoch": 1.408217097637984, "grad_norm": 0.13093435764312744, "learning_rate": 6.380162396468761e-05, "loss": 0.0272, "step": 70470 }, { "epoch": 1.4084169297789857, "grad_norm": 0.13257671892642975, "learning_rate": 6.379169098423122e-05, "loss": 0.0305, "step": 70480 }, { "epoch": 1.4086167619199872, "grad_norm": 0.1111074686050415, "learning_rate": 6.37817574146191e-05, "loss": 0.0372, "step": 70490 }, { "epoch": 1.4088165940609887, "grad_norm": 0.09779340773820877, "learning_rate": 6.377182325627562e-05, "loss": 0.0401, "step": 70500 }, { "epoch": 1.4090164262019904, "grad_norm": 0.15027756989002228, "learning_rate": 6.376188850962518e-05, "loss": 0.0306, "step": 70510 }, { "epoch": 1.409216258342992, "grad_norm": 0.1066395565867424, "learning_rate": 6.375195317509211e-05, "loss": 0.0386, "step": 70520 }, { "epoch": 1.4094160904839934, "grad_norm": 0.10489507764577866, "learning_rate": 6.374201725310089e-05, "loss": 0.0339, "step": 70530 }, { "epoch": 1.4096159226249951, "grad_norm": 0.16127993166446686, "learning_rate": 6.373208074407593e-05, "loss": 0.0342, "step": 70540 }, { "epoch": 1.4098157547659966, "grad_norm": 0.14567731320858002, "learning_rate": 6.372214364844169e-05, "loss": 0.0322, "step": 70550 }, { "epoch": 1.4100155869069981, "grad_norm": 0.1561518907546997, "learning_rate": 6.371220596662269e-05, "loss": 0.0349, "step": 70560 }, { "epoch": 1.4102154190479996, "grad_norm": 0.13044849038124084, "learning_rate": 6.370226769904346e-05, "loss": 0.0351, "step": 70570 }, { "epoch": 1.4104152511890011, "grad_norm": 0.17781288921833038, "learning_rate": 6.369232884612851e-05, "loss": 0.0358, "step": 70580 }, { "epoch": 1.4106150833300029, "grad_norm": 0.11672645062208176, "learning_rate": 6.368238940830243e-05, "loss": 0.0381, "step": 70590 }, { "epoch": 1.4108149154710043, "grad_norm": 0.14381682872772217, "learning_rate": 6.36724493859898e-05, "loss": 0.0319, "step": 70600 }, { "epoch": 1.4110147476120058, "grad_norm": 0.13139475882053375, "learning_rate": 6.366250877961525e-05, "loss": 0.0385, "step": 70610 }, { "epoch": 1.4112145797530076, "grad_norm": 0.12502406537532806, "learning_rate": 6.365256758960346e-05, "loss": 0.0352, "step": 70620 }, { "epoch": 1.411414411894009, "grad_norm": 0.12754350900650024, "learning_rate": 6.364262581637901e-05, "loss": 0.0344, "step": 70630 }, { "epoch": 1.4116142440350106, "grad_norm": 0.15685665607452393, "learning_rate": 6.363268346036669e-05, "loss": 0.03, "step": 70640 }, { "epoch": 1.411814076176012, "grad_norm": 0.11119396984577179, "learning_rate": 6.362274052199115e-05, "loss": 0.0276, "step": 70650 }, { "epoch": 1.4120139083170138, "grad_norm": 0.15978795289993286, "learning_rate": 6.361279700167717e-05, "loss": 0.0241, "step": 70660 }, { "epoch": 1.4122137404580153, "grad_norm": 0.11093681305646896, "learning_rate": 6.36028528998495e-05, "loss": 0.0304, "step": 70670 }, { "epoch": 1.4124135725990168, "grad_norm": 0.09039923548698425, "learning_rate": 6.359290821693296e-05, "loss": 0.0305, "step": 70680 }, { "epoch": 1.4126134047400183, "grad_norm": 0.13823050260543823, "learning_rate": 6.358296295335235e-05, "loss": 0.0328, "step": 70690 }, { "epoch": 1.41281323688102, "grad_norm": 0.1346491426229477, "learning_rate": 6.35730171095325e-05, "loss": 0.0298, "step": 70700 }, { "epoch": 1.4130130690220215, "grad_norm": 0.11973991245031357, "learning_rate": 6.356307068589831e-05, "loss": 0.0295, "step": 70710 }, { "epoch": 1.413212901163023, "grad_norm": 0.17224781215190887, "learning_rate": 6.355312368287464e-05, "loss": 0.0322, "step": 70720 }, { "epoch": 1.4134127333040247, "grad_norm": 0.13128070533275604, "learning_rate": 6.354317610088643e-05, "loss": 0.0353, "step": 70730 }, { "epoch": 1.4136125654450262, "grad_norm": 0.24203580617904663, "learning_rate": 6.353322794035863e-05, "loss": 0.0372, "step": 70740 }, { "epoch": 1.4138123975860277, "grad_norm": 0.221147358417511, "learning_rate": 6.352327920171618e-05, "loss": 0.0334, "step": 70750 }, { "epoch": 1.4140122297270292, "grad_norm": 0.1569819152355194, "learning_rate": 6.35133298853841e-05, "loss": 0.0348, "step": 70760 }, { "epoch": 1.4142120618680307, "grad_norm": 0.12589825689792633, "learning_rate": 6.350337999178738e-05, "loss": 0.025, "step": 70770 }, { "epoch": 1.4144118940090324, "grad_norm": 0.11799832433462143, "learning_rate": 6.349342952135107e-05, "loss": 0.0335, "step": 70780 }, { "epoch": 1.414611726150034, "grad_norm": 0.09375300258398056, "learning_rate": 6.348347847450023e-05, "loss": 0.027, "step": 70790 }, { "epoch": 1.4148115582910354, "grad_norm": 0.16490529477596283, "learning_rate": 6.347352685165998e-05, "loss": 0.0271, "step": 70800 }, { "epoch": 1.4150113904320372, "grad_norm": 0.16804902255535126, "learning_rate": 6.346357465325541e-05, "loss": 0.0298, "step": 70810 }, { "epoch": 1.4152112225730387, "grad_norm": 0.12576672434806824, "learning_rate": 6.345362187971167e-05, "loss": 0.0281, "step": 70820 }, { "epoch": 1.4154110547140402, "grad_norm": 0.10736360400915146, "learning_rate": 6.344366853145389e-05, "loss": 0.0301, "step": 70830 }, { "epoch": 1.4156108868550419, "grad_norm": 0.12436576932668686, "learning_rate": 6.343371460890733e-05, "loss": 0.0285, "step": 70840 }, { "epoch": 1.4158107189960434, "grad_norm": 0.10487372428178787, "learning_rate": 6.342376011249713e-05, "loss": 0.0277, "step": 70850 }, { "epoch": 1.4160105511370449, "grad_norm": 0.11794323474168777, "learning_rate": 6.341380504264855e-05, "loss": 0.0292, "step": 70860 }, { "epoch": 1.4162103832780464, "grad_norm": 0.11587758362293243, "learning_rate": 6.340384939978689e-05, "loss": 0.0288, "step": 70870 }, { "epoch": 1.4164102154190479, "grad_norm": 0.11228270083665848, "learning_rate": 6.33938931843374e-05, "loss": 0.0348, "step": 70880 }, { "epoch": 1.4166100475600496, "grad_norm": 0.13486696779727936, "learning_rate": 6.338393639672539e-05, "loss": 0.0338, "step": 70890 }, { "epoch": 1.416809879701051, "grad_norm": 0.11975959688425064, "learning_rate": 6.33739790373762e-05, "loss": 0.0329, "step": 70900 }, { "epoch": 1.4170097118420526, "grad_norm": 0.116387277841568, "learning_rate": 6.33640211067152e-05, "loss": 0.0282, "step": 70910 }, { "epoch": 1.4172095439830543, "grad_norm": 0.14182370901107788, "learning_rate": 6.335406260516777e-05, "loss": 0.0324, "step": 70920 }, { "epoch": 1.4174093761240558, "grad_norm": 0.11380448192358017, "learning_rate": 6.334410353315932e-05, "loss": 0.0269, "step": 70930 }, { "epoch": 1.4176092082650573, "grad_norm": 0.13925425708293915, "learning_rate": 6.333414389111528e-05, "loss": 0.0317, "step": 70940 }, { "epoch": 1.417809040406059, "grad_norm": 0.12556511163711548, "learning_rate": 6.332418367946109e-05, "loss": 0.0257, "step": 70950 }, { "epoch": 1.4180088725470605, "grad_norm": 0.1384197324514389, "learning_rate": 6.331422289862227e-05, "loss": 0.0298, "step": 70960 }, { "epoch": 1.418208704688062, "grad_norm": 0.1574431210756302, "learning_rate": 6.330426154902432e-05, "loss": 0.0278, "step": 70970 }, { "epoch": 1.4184085368290635, "grad_norm": 0.1113593727350235, "learning_rate": 6.329429963109273e-05, "loss": 0.0335, "step": 70980 }, { "epoch": 1.418608368970065, "grad_norm": 0.11545994132757187, "learning_rate": 6.328433714525311e-05, "loss": 0.0321, "step": 70990 }, { "epoch": 1.4188082011110668, "grad_norm": 0.11037024110555649, "learning_rate": 6.327437409193098e-05, "loss": 0.0324, "step": 71000 }, { "epoch": 1.4190080332520683, "grad_norm": 0.13317210972309113, "learning_rate": 6.326441047155198e-05, "loss": 0.0324, "step": 71010 }, { "epoch": 1.4192078653930698, "grad_norm": 0.11289595812559128, "learning_rate": 6.325444628454174e-05, "loss": 0.0279, "step": 71020 }, { "epoch": 1.4194076975340715, "grad_norm": 0.13760918378829956, "learning_rate": 6.324448153132593e-05, "loss": 0.0264, "step": 71030 }, { "epoch": 1.419607529675073, "grad_norm": 0.10681504756212234, "learning_rate": 6.323451621233016e-05, "loss": 0.0268, "step": 71040 }, { "epoch": 1.4198073618160745, "grad_norm": 0.14811553061008453, "learning_rate": 6.32245503279802e-05, "loss": 0.0307, "step": 71050 }, { "epoch": 1.420007193957076, "grad_norm": 0.1506650149822235, "learning_rate": 6.321458387870172e-05, "loss": 0.0286, "step": 71060 }, { "epoch": 1.4202070260980777, "grad_norm": 0.11774161458015442, "learning_rate": 6.32046168649205e-05, "loss": 0.0285, "step": 71070 }, { "epoch": 1.4204068582390792, "grad_norm": 0.110886350274086, "learning_rate": 6.319464928706231e-05, "loss": 0.0336, "step": 71080 }, { "epoch": 1.4206066903800807, "grad_norm": 0.11897682398557663, "learning_rate": 6.318468114555295e-05, "loss": 0.03, "step": 71090 }, { "epoch": 1.4208065225210822, "grad_norm": 0.1250903606414795, "learning_rate": 6.31747124408182e-05, "loss": 0.0321, "step": 71100 }, { "epoch": 1.421006354662084, "grad_norm": 0.11104627698659897, "learning_rate": 6.316474317328397e-05, "loss": 0.0299, "step": 71110 }, { "epoch": 1.4212061868030854, "grad_norm": 0.11399123817682266, "learning_rate": 6.315477334337608e-05, "loss": 0.0324, "step": 71120 }, { "epoch": 1.421406018944087, "grad_norm": 0.1336771696805954, "learning_rate": 6.314480295152045e-05, "loss": 0.025, "step": 71130 }, { "epoch": 1.4216058510850886, "grad_norm": 0.15005087852478027, "learning_rate": 6.313483199814298e-05, "loss": 0.0371, "step": 71140 }, { "epoch": 1.4218056832260901, "grad_norm": 0.10735438019037247, "learning_rate": 6.312486048366962e-05, "loss": 0.0335, "step": 71150 }, { "epoch": 1.4220055153670916, "grad_norm": 0.11519469320774078, "learning_rate": 6.311488840852633e-05, "loss": 0.0343, "step": 71160 }, { "epoch": 1.4222053475080931, "grad_norm": 0.13930074870586395, "learning_rate": 6.31049157731391e-05, "loss": 0.026, "step": 71170 }, { "epoch": 1.4224051796490946, "grad_norm": 0.11822837591171265, "learning_rate": 6.309494257793393e-05, "loss": 0.0297, "step": 71180 }, { "epoch": 1.4226050117900964, "grad_norm": 0.13348543643951416, "learning_rate": 6.308496882333687e-05, "loss": 0.0349, "step": 71190 }, { "epoch": 1.4228048439310979, "grad_norm": 0.12898892164230347, "learning_rate": 6.3074994509774e-05, "loss": 0.0308, "step": 71200 }, { "epoch": 1.4230046760720994, "grad_norm": 0.10580004006624222, "learning_rate": 6.306501963767138e-05, "loss": 0.0269, "step": 71210 }, { "epoch": 1.423204508213101, "grad_norm": 0.10331983119249344, "learning_rate": 6.30550442074551e-05, "loss": 0.0355, "step": 71220 }, { "epoch": 1.4234043403541026, "grad_norm": 0.16475969552993774, "learning_rate": 6.304506821955134e-05, "loss": 0.0356, "step": 71230 }, { "epoch": 1.423604172495104, "grad_norm": 0.09930684417486191, "learning_rate": 6.303509167438622e-05, "loss": 0.0261, "step": 71240 }, { "epoch": 1.4238040046361058, "grad_norm": 0.1071246787905693, "learning_rate": 6.302511457238592e-05, "loss": 0.0323, "step": 71250 }, { "epoch": 1.4240038367771073, "grad_norm": 0.1424228399991989, "learning_rate": 6.301513691397666e-05, "loss": 0.0285, "step": 71260 }, { "epoch": 1.4242036689181088, "grad_norm": 0.13033068180084229, "learning_rate": 6.300515869958468e-05, "loss": 0.0334, "step": 71270 }, { "epoch": 1.4244035010591103, "grad_norm": 0.1326589286327362, "learning_rate": 6.299517992963619e-05, "loss": 0.0431, "step": 71280 }, { "epoch": 1.4246033332001118, "grad_norm": 0.13788406550884247, "learning_rate": 6.29852006045575e-05, "loss": 0.0336, "step": 71290 }, { "epoch": 1.4248031653411135, "grad_norm": 0.1248350590467453, "learning_rate": 6.297522072477487e-05, "loss": 0.0313, "step": 71300 }, { "epoch": 1.425002997482115, "grad_norm": 0.13306276500225067, "learning_rate": 6.296524029071467e-05, "loss": 0.0307, "step": 71310 }, { "epoch": 1.4252028296231165, "grad_norm": 0.12235765904188156, "learning_rate": 6.295525930280321e-05, "loss": 0.0257, "step": 71320 }, { "epoch": 1.4254026617641182, "grad_norm": 0.16273066401481628, "learning_rate": 6.294527776146688e-05, "loss": 0.0327, "step": 71330 }, { "epoch": 1.4256024939051197, "grad_norm": 0.13205349445343018, "learning_rate": 6.293529566713207e-05, "loss": 0.038, "step": 71340 }, { "epoch": 1.4258023260461212, "grad_norm": 0.11581528931856155, "learning_rate": 6.292531302022518e-05, "loss": 0.0314, "step": 71350 }, { "epoch": 1.426002158187123, "grad_norm": 0.10937788337469101, "learning_rate": 6.291532982117267e-05, "loss": 0.0285, "step": 71360 }, { "epoch": 1.4262019903281244, "grad_norm": 0.1410306692123413, "learning_rate": 6.2905346070401e-05, "loss": 0.0357, "step": 71370 }, { "epoch": 1.426401822469126, "grad_norm": 0.12900973856449127, "learning_rate": 6.289536176833664e-05, "loss": 0.0299, "step": 71380 }, { "epoch": 1.4266016546101274, "grad_norm": 0.13355118036270142, "learning_rate": 6.288537691540612e-05, "loss": 0.0368, "step": 71390 }, { "epoch": 1.426801486751129, "grad_norm": 0.13654476404190063, "learning_rate": 6.287539151203598e-05, "loss": 0.0354, "step": 71400 }, { "epoch": 1.4270013188921307, "grad_norm": 0.12863852083683014, "learning_rate": 6.286540555865277e-05, "loss": 0.0302, "step": 71410 }, { "epoch": 1.4272011510331322, "grad_norm": 0.14660191535949707, "learning_rate": 6.285541905568304e-05, "loss": 0.0311, "step": 71420 }, { "epoch": 1.4274009831741337, "grad_norm": 0.11491204798221588, "learning_rate": 6.284543200355345e-05, "loss": 0.0313, "step": 71430 }, { "epoch": 1.4276008153151354, "grad_norm": 0.1396121382713318, "learning_rate": 6.283544440269061e-05, "loss": 0.033, "step": 71440 }, { "epoch": 1.4278006474561369, "grad_norm": 0.1272425353527069, "learning_rate": 6.282545625352112e-05, "loss": 0.0359, "step": 71450 }, { "epoch": 1.4280004795971384, "grad_norm": 0.14112116396427155, "learning_rate": 6.281546755647174e-05, "loss": 0.0409, "step": 71460 }, { "epoch": 1.4282003117381399, "grad_norm": 0.1271960437297821, "learning_rate": 6.280547831196913e-05, "loss": 0.0294, "step": 71470 }, { "epoch": 1.4284001438791416, "grad_norm": 0.12172341346740723, "learning_rate": 6.279548852044e-05, "loss": 0.027, "step": 71480 }, { "epoch": 1.428599976020143, "grad_norm": 0.1438077837228775, "learning_rate": 6.27854981823111e-05, "loss": 0.0252, "step": 71490 }, { "epoch": 1.4287998081611446, "grad_norm": 0.13529834151268005, "learning_rate": 6.27755072980092e-05, "loss": 0.0282, "step": 71500 }, { "epoch": 1.428999640302146, "grad_norm": 0.14151713252067566, "learning_rate": 6.276551586796113e-05, "loss": 0.029, "step": 71510 }, { "epoch": 1.4291994724431478, "grad_norm": 0.13511653244495392, "learning_rate": 6.275552389259363e-05, "loss": 0.035, "step": 71520 }, { "epoch": 1.4293993045841493, "grad_norm": 0.12885195016860962, "learning_rate": 6.274553137233361e-05, "loss": 0.0302, "step": 71530 }, { "epoch": 1.4295991367251508, "grad_norm": 0.1406830996274948, "learning_rate": 6.273553830760791e-05, "loss": 0.0319, "step": 71540 }, { "epoch": 1.4297989688661525, "grad_norm": 0.10298257321119308, "learning_rate": 6.272554469884338e-05, "loss": 0.0307, "step": 71550 }, { "epoch": 1.429998801007154, "grad_norm": 0.13546036183834076, "learning_rate": 6.271555054646699e-05, "loss": 0.0272, "step": 71560 }, { "epoch": 1.4301986331481555, "grad_norm": 0.142604798078537, "learning_rate": 6.270555585090563e-05, "loss": 0.027, "step": 71570 }, { "epoch": 1.430398465289157, "grad_norm": 0.12914209067821503, "learning_rate": 6.269556061258626e-05, "loss": 0.034, "step": 71580 }, { "epoch": 1.4305982974301585, "grad_norm": 0.1623460054397583, "learning_rate": 6.268556483193586e-05, "loss": 0.0216, "step": 71590 }, { "epoch": 1.4307981295711603, "grad_norm": 0.1667923778295517, "learning_rate": 6.267556850938144e-05, "loss": 0.0367, "step": 71600 }, { "epoch": 1.4309979617121618, "grad_norm": 0.1538841277360916, "learning_rate": 6.266557164535002e-05, "loss": 0.032, "step": 71610 }, { "epoch": 1.4311977938531633, "grad_norm": 0.1286778599023819, "learning_rate": 6.265557424026862e-05, "loss": 0.0325, "step": 71620 }, { "epoch": 1.431397625994165, "grad_norm": 0.15415059030056, "learning_rate": 6.264557629456438e-05, "loss": 0.0288, "step": 71630 }, { "epoch": 1.4315974581351665, "grad_norm": 0.14757023751735687, "learning_rate": 6.263557780866433e-05, "loss": 0.0293, "step": 71640 }, { "epoch": 1.431797290276168, "grad_norm": 0.16256588697433472, "learning_rate": 6.262557878299562e-05, "loss": 0.035, "step": 71650 }, { "epoch": 1.4319971224171697, "grad_norm": 0.12293776124715805, "learning_rate": 6.261557921798535e-05, "loss": 0.0382, "step": 71660 }, { "epoch": 1.4321969545581712, "grad_norm": 0.12390564382076263, "learning_rate": 6.260557911406073e-05, "loss": 0.0314, "step": 71670 }, { "epoch": 1.4323967866991727, "grad_norm": 0.16451995074748993, "learning_rate": 6.259557847164894e-05, "loss": 0.0343, "step": 71680 }, { "epoch": 1.4325966188401742, "grad_norm": 0.12976469099521637, "learning_rate": 6.258557729117717e-05, "loss": 0.0297, "step": 71690 }, { "epoch": 1.4327964509811757, "grad_norm": 0.11956112086772919, "learning_rate": 6.257557557307265e-05, "loss": 0.0326, "step": 71700 }, { "epoch": 1.4329962831221774, "grad_norm": 0.1367029845714569, "learning_rate": 6.256557331776264e-05, "loss": 0.0276, "step": 71710 }, { "epoch": 1.433196115263179, "grad_norm": 0.11696581542491913, "learning_rate": 6.255557052567442e-05, "loss": 0.0286, "step": 71720 }, { "epoch": 1.4333959474041804, "grad_norm": 0.11986507475376129, "learning_rate": 6.25455671972353e-05, "loss": 0.0341, "step": 71730 }, { "epoch": 1.4335957795451821, "grad_norm": 0.14241942763328552, "learning_rate": 6.25355633328726e-05, "loss": 0.0302, "step": 71740 }, { "epoch": 1.4337956116861836, "grad_norm": 0.12422201782464981, "learning_rate": 6.252555893301366e-05, "loss": 0.0311, "step": 71750 }, { "epoch": 1.4339954438271851, "grad_norm": 0.11873544752597809, "learning_rate": 6.251555399808586e-05, "loss": 0.0309, "step": 71760 }, { "epoch": 1.4341952759681869, "grad_norm": 0.1536053717136383, "learning_rate": 6.250554852851658e-05, "loss": 0.0346, "step": 71770 }, { "epoch": 1.4343951081091884, "grad_norm": 0.1056247279047966, "learning_rate": 6.249554252473323e-05, "loss": 0.0209, "step": 71780 }, { "epoch": 1.4345949402501899, "grad_norm": 0.16370977461338043, "learning_rate": 6.248553598716329e-05, "loss": 0.0304, "step": 71790 }, { "epoch": 1.4347947723911914, "grad_norm": 0.13253246247768402, "learning_rate": 6.247552891623413e-05, "loss": 0.0366, "step": 71800 }, { "epoch": 1.4349946045321929, "grad_norm": 0.11685065925121307, "learning_rate": 6.246552131237335e-05, "loss": 0.0295, "step": 71810 }, { "epoch": 1.4351944366731946, "grad_norm": 0.11969586461782455, "learning_rate": 6.245551317600836e-05, "loss": 0.0291, "step": 71820 }, { "epoch": 1.435394268814196, "grad_norm": 0.10834675282239914, "learning_rate": 6.244550450756675e-05, "loss": 0.027, "step": 71830 }, { "epoch": 1.4355941009551976, "grad_norm": 0.122909314930439, "learning_rate": 6.243549530747606e-05, "loss": 0.0345, "step": 71840 }, { "epoch": 1.4357939330961993, "grad_norm": 0.13285775482654572, "learning_rate": 6.242548557616383e-05, "loss": 0.0383, "step": 71850 }, { "epoch": 1.4359937652372008, "grad_norm": 0.14536045491695404, "learning_rate": 6.241547531405769e-05, "loss": 0.0346, "step": 71860 }, { "epoch": 1.4361935973782023, "grad_norm": 0.1304694265127182, "learning_rate": 6.240546452158527e-05, "loss": 0.0305, "step": 71870 }, { "epoch": 1.4363934295192038, "grad_norm": 0.11943697184324265, "learning_rate": 6.239545319917418e-05, "loss": 0.0292, "step": 71880 }, { "epoch": 1.4365932616602053, "grad_norm": 0.1960320919752121, "learning_rate": 6.23854413472521e-05, "loss": 0.0297, "step": 71890 }, { "epoch": 1.436793093801207, "grad_norm": 0.1322428286075592, "learning_rate": 6.237542896624673e-05, "loss": 0.0272, "step": 71900 }, { "epoch": 1.4369929259422085, "grad_norm": 0.12728658318519592, "learning_rate": 6.236541605658575e-05, "loss": 0.023, "step": 71910 }, { "epoch": 1.43719275808321, "grad_norm": 0.11654120683670044, "learning_rate": 6.235540261869692e-05, "loss": 0.0294, "step": 71920 }, { "epoch": 1.4373925902242117, "grad_norm": 0.11138120293617249, "learning_rate": 6.2345388653008e-05, "loss": 0.0219, "step": 71930 }, { "epoch": 1.4375924223652132, "grad_norm": 0.11055946350097656, "learning_rate": 6.233537415994675e-05, "loss": 0.0338, "step": 71940 }, { "epoch": 1.4377922545062147, "grad_norm": 0.11408444494009018, "learning_rate": 6.232535913994097e-05, "loss": 0.0333, "step": 71950 }, { "epoch": 1.4379920866472164, "grad_norm": 0.12110147625207901, "learning_rate": 6.231534359341849e-05, "loss": 0.0305, "step": 71960 }, { "epoch": 1.438191918788218, "grad_norm": 0.13268250226974487, "learning_rate": 6.230532752080715e-05, "loss": 0.0285, "step": 71970 }, { "epoch": 1.4383917509292194, "grad_norm": 0.11136294156312943, "learning_rate": 6.229531092253484e-05, "loss": 0.0311, "step": 71980 }, { "epoch": 1.438591583070221, "grad_norm": 0.12537835538387299, "learning_rate": 6.228529379902943e-05, "loss": 0.0338, "step": 71990 }, { "epoch": 1.4387914152112224, "grad_norm": 0.17029838263988495, "learning_rate": 6.227527615071882e-05, "loss": 0.0293, "step": 72000 }, { "epoch": 1.4389912473522242, "grad_norm": 0.11274378001689911, "learning_rate": 6.2265257978031e-05, "loss": 0.0279, "step": 72010 }, { "epoch": 1.4391910794932257, "grad_norm": 0.09400060027837753, "learning_rate": 6.225523928139385e-05, "loss": 0.0343, "step": 72020 }, { "epoch": 1.4393909116342272, "grad_norm": 0.167488694190979, "learning_rate": 6.224522006123543e-05, "loss": 0.0261, "step": 72030 }, { "epoch": 1.4395907437752289, "grad_norm": 0.11726343631744385, "learning_rate": 6.223520031798368e-05, "loss": 0.0311, "step": 72040 }, { "epoch": 1.4397905759162304, "grad_norm": 0.10376575589179993, "learning_rate": 6.222518005206665e-05, "loss": 0.0336, "step": 72050 }, { "epoch": 1.4399904080572319, "grad_norm": 0.11006120592355728, "learning_rate": 6.221515926391239e-05, "loss": 0.0314, "step": 72060 }, { "epoch": 1.4401902401982336, "grad_norm": 0.12548735737800598, "learning_rate": 6.220513795394896e-05, "loss": 0.0305, "step": 72070 }, { "epoch": 1.440390072339235, "grad_norm": 0.12904953956604004, "learning_rate": 6.219511612260446e-05, "loss": 0.0271, "step": 72080 }, { "epoch": 1.4405899044802366, "grad_norm": 0.14558197557926178, "learning_rate": 6.218509377030702e-05, "loss": 0.0299, "step": 72090 }, { "epoch": 1.440789736621238, "grad_norm": 0.13174481689929962, "learning_rate": 6.217507089748475e-05, "loss": 0.0328, "step": 72100 }, { "epoch": 1.4409895687622396, "grad_norm": 0.11892367154359818, "learning_rate": 6.21650475045658e-05, "loss": 0.0279, "step": 72110 }, { "epoch": 1.4411894009032413, "grad_norm": 0.10165022313594818, "learning_rate": 6.215502359197839e-05, "loss": 0.0384, "step": 72120 }, { "epoch": 1.4413892330442428, "grad_norm": 0.1357623040676117, "learning_rate": 6.214499916015069e-05, "loss": 0.0364, "step": 72130 }, { "epoch": 1.4415890651852443, "grad_norm": 0.12734471261501312, "learning_rate": 6.213497420951095e-05, "loss": 0.0305, "step": 72140 }, { "epoch": 1.441788897326246, "grad_norm": 0.10792866349220276, "learning_rate": 6.21249487404874e-05, "loss": 0.0279, "step": 72150 }, { "epoch": 1.4419887294672475, "grad_norm": 0.12534010410308838, "learning_rate": 6.21149227535083e-05, "loss": 0.0344, "step": 72160 }, { "epoch": 1.442188561608249, "grad_norm": 0.1458064764738083, "learning_rate": 6.210489624900196e-05, "loss": 0.0322, "step": 72170 }, { "epoch": 1.4423883937492508, "grad_norm": 0.13635283708572388, "learning_rate": 6.20948692273967e-05, "loss": 0.0409, "step": 72180 }, { "epoch": 1.4425882258902523, "grad_norm": 0.12452627718448639, "learning_rate": 6.208484168912084e-05, "loss": 0.0333, "step": 72190 }, { "epoch": 1.4427880580312538, "grad_norm": 0.14993397891521454, "learning_rate": 6.207481363460274e-05, "loss": 0.0233, "step": 72200 }, { "epoch": 1.4429878901722553, "grad_norm": 0.14113470911979675, "learning_rate": 6.206478506427079e-05, "loss": 0.0319, "step": 72210 }, { "epoch": 1.4431877223132568, "grad_norm": 0.1365744173526764, "learning_rate": 6.205475597855338e-05, "loss": 0.0296, "step": 72220 }, { "epoch": 1.4433875544542585, "grad_norm": 0.12035641819238663, "learning_rate": 6.204472637787895e-05, "loss": 0.0254, "step": 72230 }, { "epoch": 1.44358738659526, "grad_norm": 0.1304968148469925, "learning_rate": 6.203469626267593e-05, "loss": 0.028, "step": 72240 }, { "epoch": 1.4437872187362615, "grad_norm": 0.11172423511743546, "learning_rate": 6.202466563337279e-05, "loss": 0.037, "step": 72250 }, { "epoch": 1.4439870508772632, "grad_norm": 0.1189531534910202, "learning_rate": 6.201463449039804e-05, "loss": 0.029, "step": 72260 }, { "epoch": 1.4441868830182647, "grad_norm": 0.13310667872428894, "learning_rate": 6.200460283418014e-05, "loss": 0.0364, "step": 72270 }, { "epoch": 1.4443867151592662, "grad_norm": 0.15300524234771729, "learning_rate": 6.19945706651477e-05, "loss": 0.0294, "step": 72280 }, { "epoch": 1.4445865473002677, "grad_norm": 0.1923244446516037, "learning_rate": 6.198453798372922e-05, "loss": 0.0321, "step": 72290 }, { "epoch": 1.4447863794412692, "grad_norm": 0.12080588191747665, "learning_rate": 6.197450479035329e-05, "loss": 0.0326, "step": 72300 }, { "epoch": 1.444986211582271, "grad_norm": 0.11698614060878754, "learning_rate": 6.196447108544851e-05, "loss": 0.0295, "step": 72310 }, { "epoch": 1.4451860437232724, "grad_norm": 0.11549145728349686, "learning_rate": 6.195443686944352e-05, "loss": 0.0312, "step": 72320 }, { "epoch": 1.445385875864274, "grad_norm": 0.11432341486215591, "learning_rate": 6.194440214276694e-05, "loss": 0.0298, "step": 72330 }, { "epoch": 1.4455857080052756, "grad_norm": 0.11751812696456909, "learning_rate": 6.193436690584744e-05, "loss": 0.0276, "step": 72340 }, { "epoch": 1.4457855401462771, "grad_norm": 0.13824176788330078, "learning_rate": 6.192433115911371e-05, "loss": 0.0335, "step": 72350 }, { "epoch": 1.4459853722872786, "grad_norm": 0.14488773047924042, "learning_rate": 6.191429490299446e-05, "loss": 0.0368, "step": 72360 }, { "epoch": 1.4461852044282804, "grad_norm": 0.1292170286178589, "learning_rate": 6.190425813791843e-05, "loss": 0.0293, "step": 72370 }, { "epoch": 1.4463850365692819, "grad_norm": 0.10921189934015274, "learning_rate": 6.189422086431433e-05, "loss": 0.0297, "step": 72380 }, { "epoch": 1.4465848687102834, "grad_norm": 0.12706691026687622, "learning_rate": 6.188418308261101e-05, "loss": 0.0279, "step": 72390 }, { "epoch": 1.4467847008512849, "grad_norm": 0.210713729262352, "learning_rate": 6.187414479323717e-05, "loss": 0.0256, "step": 72400 }, { "epoch": 1.4469845329922864, "grad_norm": 0.11228211969137192, "learning_rate": 6.186410599662172e-05, "loss": 0.0298, "step": 72410 }, { "epoch": 1.447184365133288, "grad_norm": 0.15068365633487701, "learning_rate": 6.185406669319343e-05, "loss": 0.0324, "step": 72420 }, { "epoch": 1.4473841972742896, "grad_norm": 0.12799689173698425, "learning_rate": 6.184402688338121e-05, "loss": 0.0239, "step": 72430 }, { "epoch": 1.447584029415291, "grad_norm": 0.14700539410114288, "learning_rate": 6.183398656761391e-05, "loss": 0.0343, "step": 72440 }, { "epoch": 1.4477838615562928, "grad_norm": 0.11491595208644867, "learning_rate": 6.182394574632043e-05, "loss": 0.0268, "step": 72450 }, { "epoch": 1.4479836936972943, "grad_norm": 0.12261638045310974, "learning_rate": 6.181390441992974e-05, "loss": 0.0292, "step": 72460 }, { "epoch": 1.4481835258382958, "grad_norm": 0.15615054965019226, "learning_rate": 6.180386258887072e-05, "loss": 0.0344, "step": 72470 }, { "epoch": 1.4483833579792975, "grad_norm": 0.1346363127231598, "learning_rate": 6.17938202535724e-05, "loss": 0.0321, "step": 72480 }, { "epoch": 1.448583190120299, "grad_norm": 0.14489831030368805, "learning_rate": 6.178377741446374e-05, "loss": 0.0261, "step": 72490 }, { "epoch": 1.4487830222613005, "grad_norm": 0.1204143539071083, "learning_rate": 6.177373407197376e-05, "loss": 0.0222, "step": 72500 }, { "epoch": 1.448982854402302, "grad_norm": 0.13464288413524628, "learning_rate": 6.176369022653147e-05, "loss": 0.0325, "step": 72510 }, { "epoch": 1.4491826865433035, "grad_norm": 0.1278420239686966, "learning_rate": 6.175364587856595e-05, "loss": 0.0376, "step": 72520 }, { "epoch": 1.4493825186843052, "grad_norm": 0.1806410402059555, "learning_rate": 6.17436010285063e-05, "loss": 0.0353, "step": 72530 }, { "epoch": 1.4495823508253067, "grad_norm": 0.14912568032741547, "learning_rate": 6.173355567678155e-05, "loss": 0.0393, "step": 72540 }, { "epoch": 1.4497821829663082, "grad_norm": 0.13440929353237152, "learning_rate": 6.172350982382087e-05, "loss": 0.037, "step": 72550 }, { "epoch": 1.44998201510731, "grad_norm": 0.126108318567276, "learning_rate": 6.171346347005339e-05, "loss": 0.0453, "step": 72560 }, { "epoch": 1.4501818472483115, "grad_norm": 0.1409144401550293, "learning_rate": 6.170341661590826e-05, "loss": 0.0264, "step": 72570 }, { "epoch": 1.450381679389313, "grad_norm": 0.16485930979251862, "learning_rate": 6.169336926181468e-05, "loss": 0.0305, "step": 72580 }, { "epoch": 1.4505815115303147, "grad_norm": 0.12566953897476196, "learning_rate": 6.168332140820185e-05, "loss": 0.0309, "step": 72590 }, { "epoch": 1.4507813436713162, "grad_norm": 0.16733422875404358, "learning_rate": 6.167327305549898e-05, "loss": 0.0266, "step": 72600 }, { "epoch": 1.4509811758123177, "grad_norm": 0.12787692248821259, "learning_rate": 6.166322420413536e-05, "loss": 0.0346, "step": 72610 }, { "epoch": 1.4511810079533192, "grad_norm": 0.19697560369968414, "learning_rate": 6.165317485454019e-05, "loss": 0.028, "step": 72620 }, { "epoch": 1.4513808400943207, "grad_norm": 0.1313786506652832, "learning_rate": 6.164312500714281e-05, "loss": 0.0315, "step": 72630 }, { "epoch": 1.4515806722353224, "grad_norm": 0.12000704556703568, "learning_rate": 6.163307466237253e-05, "loss": 0.0274, "step": 72640 }, { "epoch": 1.4517805043763239, "grad_norm": 0.1456483155488968, "learning_rate": 6.162302382065865e-05, "loss": 0.0368, "step": 72650 }, { "epoch": 1.4519803365173254, "grad_norm": 0.1263597458600998, "learning_rate": 6.161297248243058e-05, "loss": 0.0331, "step": 72660 }, { "epoch": 1.452180168658327, "grad_norm": 0.13658507168293, "learning_rate": 6.160292064811761e-05, "loss": 0.0314, "step": 72670 }, { "epoch": 1.4523800007993286, "grad_norm": 0.12759090960025787, "learning_rate": 6.159286831814924e-05, "loss": 0.0298, "step": 72680 }, { "epoch": 1.45257983294033, "grad_norm": 0.16263249516487122, "learning_rate": 6.158281549295479e-05, "loss": 0.0365, "step": 72690 }, { "epoch": 1.4527796650813316, "grad_norm": 0.12009211629629135, "learning_rate": 6.157276217296375e-05, "loss": 0.0301, "step": 72700 }, { "epoch": 1.452979497222333, "grad_norm": 0.10871230065822601, "learning_rate": 6.15627083586056e-05, "loss": 0.0329, "step": 72710 }, { "epoch": 1.4531793293633348, "grad_norm": 0.15269744396209717, "learning_rate": 6.155265405030974e-05, "loss": 0.0241, "step": 72720 }, { "epoch": 1.4533791615043363, "grad_norm": 0.11973373591899872, "learning_rate": 6.154259924850576e-05, "loss": 0.0342, "step": 72730 }, { "epoch": 1.4535789936453378, "grad_norm": 0.1435864418745041, "learning_rate": 6.153254395362312e-05, "loss": 0.0338, "step": 72740 }, { "epoch": 1.4537788257863395, "grad_norm": 0.22207851707935333, "learning_rate": 6.152248816609141e-05, "loss": 0.0305, "step": 72750 }, { "epoch": 1.453978657927341, "grad_norm": 0.14011913537979126, "learning_rate": 6.151243188634016e-05, "loss": 0.0323, "step": 72760 }, { "epoch": 1.4541784900683425, "grad_norm": 0.1600448340177536, "learning_rate": 6.150237511479897e-05, "loss": 0.0321, "step": 72770 }, { "epoch": 1.4543783222093443, "grad_norm": 0.12183921784162521, "learning_rate": 6.149231785189744e-05, "loss": 0.0292, "step": 72780 }, { "epoch": 1.4545781543503458, "grad_norm": 0.14995858073234558, "learning_rate": 6.148226009806523e-05, "loss": 0.0248, "step": 72790 }, { "epoch": 1.4547779864913473, "grad_norm": 0.12457888573408127, "learning_rate": 6.147220185373194e-05, "loss": 0.028, "step": 72800 }, { "epoch": 1.4549778186323488, "grad_norm": 0.12175686657428741, "learning_rate": 6.146214311932725e-05, "loss": 0.0327, "step": 72810 }, { "epoch": 1.4551776507733503, "grad_norm": 0.14778754115104675, "learning_rate": 6.145208389528088e-05, "loss": 0.0282, "step": 72820 }, { "epoch": 1.455377482914352, "grad_norm": 0.1274305284023285, "learning_rate": 6.144202418202251e-05, "loss": 0.0304, "step": 72830 }, { "epoch": 1.4555773150553535, "grad_norm": 0.1136329397559166, "learning_rate": 6.143196397998191e-05, "loss": 0.0237, "step": 72840 }, { "epoch": 1.455777147196355, "grad_norm": 0.16296035051345825, "learning_rate": 6.142190328958879e-05, "loss": 0.0407, "step": 72850 }, { "epoch": 1.4559769793373567, "grad_norm": 0.20133286714553833, "learning_rate": 6.141184211127297e-05, "loss": 0.0381, "step": 72860 }, { "epoch": 1.4561768114783582, "grad_norm": 0.14977452158927917, "learning_rate": 6.140178044546419e-05, "loss": 0.0307, "step": 72870 }, { "epoch": 1.4563766436193597, "grad_norm": 0.21678133308887482, "learning_rate": 6.139171829259231e-05, "loss": 0.0337, "step": 72880 }, { "epoch": 1.4565764757603614, "grad_norm": 0.12786458432674408, "learning_rate": 6.138165565308715e-05, "loss": 0.0315, "step": 72890 }, { "epoch": 1.456776307901363, "grad_norm": 0.1216253936290741, "learning_rate": 6.137159252737857e-05, "loss": 0.0252, "step": 72900 }, { "epoch": 1.4569761400423644, "grad_norm": 0.08689599484205246, "learning_rate": 6.136152891589644e-05, "loss": 0.0307, "step": 72910 }, { "epoch": 1.457175972183366, "grad_norm": 0.13692113757133484, "learning_rate": 6.135146481907067e-05, "loss": 0.0272, "step": 72920 }, { "epoch": 1.4573758043243674, "grad_norm": 0.1361297369003296, "learning_rate": 6.134140023733119e-05, "loss": 0.036, "step": 72930 }, { "epoch": 1.4575756364653691, "grad_norm": 0.12400926649570465, "learning_rate": 6.133133517110791e-05, "loss": 0.0284, "step": 72940 }, { "epoch": 1.4577754686063706, "grad_norm": 0.1413065493106842, "learning_rate": 6.132126962083081e-05, "loss": 0.0303, "step": 72950 }, { "epoch": 1.4579753007473721, "grad_norm": 0.1251256912946701, "learning_rate": 6.131120358692986e-05, "loss": 0.0296, "step": 72960 }, { "epoch": 1.4581751328883739, "grad_norm": 0.14360588788986206, "learning_rate": 6.130113706983508e-05, "loss": 0.0282, "step": 72970 }, { "epoch": 1.4583749650293754, "grad_norm": 0.3666026294231415, "learning_rate": 6.12910700699765e-05, "loss": 0.0391, "step": 72980 }, { "epoch": 1.4585747971703769, "grad_norm": 0.13145416975021362, "learning_rate": 6.128100258778414e-05, "loss": 0.0278, "step": 72990 }, { "epoch": 1.4587746293113786, "grad_norm": 0.16006450355052948, "learning_rate": 6.127093462368807e-05, "loss": 0.0339, "step": 73000 }, { "epoch": 1.45897446145238, "grad_norm": 0.1322186440229416, "learning_rate": 6.126086617811839e-05, "loss": 0.0319, "step": 73010 }, { "epoch": 1.4591742935933816, "grad_norm": 0.13593463599681854, "learning_rate": 6.125079725150518e-05, "loss": 0.0371, "step": 73020 }, { "epoch": 1.459374125734383, "grad_norm": 0.14975622296333313, "learning_rate": 6.124072784427858e-05, "loss": 0.0316, "step": 73030 }, { "epoch": 1.4595739578753846, "grad_norm": 0.10127629339694977, "learning_rate": 6.123065795686872e-05, "loss": 0.0272, "step": 73040 }, { "epoch": 1.4597737900163863, "grad_norm": 0.14875145256519318, "learning_rate": 6.122058758970581e-05, "loss": 0.0296, "step": 73050 }, { "epoch": 1.4599736221573878, "grad_norm": 0.14250363409519196, "learning_rate": 6.121051674322002e-05, "loss": 0.0343, "step": 73060 }, { "epoch": 1.4601734542983893, "grad_norm": 0.1350213885307312, "learning_rate": 6.120044541784152e-05, "loss": 0.0407, "step": 73070 }, { "epoch": 1.460373286439391, "grad_norm": 0.1430635303258896, "learning_rate": 6.119037361400059e-05, "loss": 0.0273, "step": 73080 }, { "epoch": 1.4605731185803925, "grad_norm": 0.1260179728269577, "learning_rate": 6.118030133212744e-05, "loss": 0.0299, "step": 73090 }, { "epoch": 1.460772950721394, "grad_norm": 0.15494507551193237, "learning_rate": 6.117022857265237e-05, "loss": 0.0351, "step": 73100 }, { "epoch": 1.4609727828623955, "grad_norm": 0.1567942500114441, "learning_rate": 6.116015533600566e-05, "loss": 0.031, "step": 73110 }, { "epoch": 1.461172615003397, "grad_norm": 0.12324915081262589, "learning_rate": 6.11500816226176e-05, "loss": 0.0291, "step": 73120 }, { "epoch": 1.4613724471443987, "grad_norm": 0.1361856609582901, "learning_rate": 6.114000743291854e-05, "loss": 0.0344, "step": 73130 }, { "epoch": 1.4615722792854002, "grad_norm": 0.16405999660491943, "learning_rate": 6.112993276733882e-05, "loss": 0.0311, "step": 73140 }, { "epoch": 1.4617721114264017, "grad_norm": 0.11854039132595062, "learning_rate": 6.111985762630883e-05, "loss": 0.0269, "step": 73150 }, { "epoch": 1.4619719435674035, "grad_norm": 0.1521015614271164, "learning_rate": 6.110978201025896e-05, "loss": 0.036, "step": 73160 }, { "epoch": 1.462171775708405, "grad_norm": 0.16255873441696167, "learning_rate": 6.109970591961959e-05, "loss": 0.0355, "step": 73170 }, { "epoch": 1.4623716078494065, "grad_norm": 0.1474950909614563, "learning_rate": 6.108962935482118e-05, "loss": 0.0343, "step": 73180 }, { "epoch": 1.4625714399904082, "grad_norm": 0.12041549384593964, "learning_rate": 6.107955231629418e-05, "loss": 0.0323, "step": 73190 }, { "epoch": 1.4627712721314097, "grad_norm": 0.14579829573631287, "learning_rate": 6.106947480446906e-05, "loss": 0.0233, "step": 73200 }, { "epoch": 1.4629711042724112, "grad_norm": 0.10527361929416656, "learning_rate": 6.105939681977631e-05, "loss": 0.0325, "step": 73210 }, { "epoch": 1.4631709364134127, "grad_norm": 0.1383107304573059, "learning_rate": 6.104931836264644e-05, "loss": 0.0281, "step": 73220 }, { "epoch": 1.4633707685544142, "grad_norm": 0.10500621050596237, "learning_rate": 6.103923943351e-05, "loss": 0.0386, "step": 73230 }, { "epoch": 1.463570600695416, "grad_norm": 0.19009873270988464, "learning_rate": 6.1029160032797514e-05, "loss": 0.0383, "step": 73240 }, { "epoch": 1.4637704328364174, "grad_norm": 0.13518434762954712, "learning_rate": 6.101908016093959e-05, "loss": 0.0244, "step": 73250 }, { "epoch": 1.463970264977419, "grad_norm": 0.1283038854598999, "learning_rate": 6.1008999818366807e-05, "loss": 0.0299, "step": 73260 }, { "epoch": 1.4641700971184206, "grad_norm": 0.10580570995807648, "learning_rate": 6.099891900550976e-05, "loss": 0.031, "step": 73270 }, { "epoch": 1.464369929259422, "grad_norm": 0.10066284239292145, "learning_rate": 6.0988837722799095e-05, "loss": 0.0308, "step": 73280 }, { "epoch": 1.4645697614004236, "grad_norm": 0.1462107002735138, "learning_rate": 6.097875597066548e-05, "loss": 0.0351, "step": 73290 }, { "epoch": 1.4647695935414253, "grad_norm": 0.12009968608617783, "learning_rate": 6.096867374953958e-05, "loss": 0.0354, "step": 73300 }, { "epoch": 1.4649694256824268, "grad_norm": 0.15207096934318542, "learning_rate": 6.095859105985209e-05, "loss": 0.0323, "step": 73310 }, { "epoch": 1.4651692578234283, "grad_norm": 0.15150968730449677, "learning_rate": 6.0948507902033704e-05, "loss": 0.039, "step": 73320 }, { "epoch": 1.4653690899644298, "grad_norm": 0.17038773000240326, "learning_rate": 6.093842427651519e-05, "loss": 0.028, "step": 73330 }, { "epoch": 1.4655689221054313, "grad_norm": 0.12893545627593994, "learning_rate": 6.092834018372728e-05, "loss": 0.0301, "step": 73340 }, { "epoch": 1.465768754246433, "grad_norm": 0.13685929775238037, "learning_rate": 6.091825562410076e-05, "loss": 0.0396, "step": 73350 }, { "epoch": 1.4659685863874345, "grad_norm": 0.1303589940071106, "learning_rate": 6.0908170598066415e-05, "loss": 0.0297, "step": 73360 }, { "epoch": 1.466168418528436, "grad_norm": 0.11243166774511337, "learning_rate": 6.089808510605505e-05, "loss": 0.0281, "step": 73370 }, { "epoch": 1.4663682506694378, "grad_norm": 0.12501762807369232, "learning_rate": 6.0887999148497535e-05, "loss": 0.0323, "step": 73380 }, { "epoch": 1.4665680828104393, "grad_norm": 0.1362762451171875, "learning_rate": 6.087791272582466e-05, "loss": 0.0297, "step": 73390 }, { "epoch": 1.4667679149514408, "grad_norm": 0.18362146615982056, "learning_rate": 6.0867825838467366e-05, "loss": 0.0285, "step": 73400 }, { "epoch": 1.4669677470924425, "grad_norm": 0.1467016637325287, "learning_rate": 6.085773848685651e-05, "loss": 0.0318, "step": 73410 }, { "epoch": 1.467167579233444, "grad_norm": 0.13383494317531586, "learning_rate": 6.084765067142301e-05, "loss": 0.0379, "step": 73420 }, { "epoch": 1.4673674113744455, "grad_norm": 0.12178131937980652, "learning_rate": 6.083756239259779e-05, "loss": 0.0225, "step": 73430 }, { "epoch": 1.467567243515447, "grad_norm": 0.14695388078689575, "learning_rate": 6.082747365081183e-05, "loss": 0.0315, "step": 73440 }, { "epoch": 1.4677670756564485, "grad_norm": 0.12541928887367249, "learning_rate": 6.081738444649609e-05, "loss": 0.0325, "step": 73450 }, { "epoch": 1.4679669077974502, "grad_norm": 0.13348379731178284, "learning_rate": 6.0807294780081545e-05, "loss": 0.0384, "step": 73460 }, { "epoch": 1.4681667399384517, "grad_norm": 0.12071783095598221, "learning_rate": 6.0797204651999226e-05, "loss": 0.0335, "step": 73470 }, { "epoch": 1.4683665720794532, "grad_norm": 0.17925646901130676, "learning_rate": 6.0787114062680165e-05, "loss": 0.0299, "step": 73480 }, { "epoch": 1.468566404220455, "grad_norm": 0.1255030333995819, "learning_rate": 6.07770230125554e-05, "loss": 0.0395, "step": 73490 }, { "epoch": 1.4687662363614564, "grad_norm": 0.14637427031993866, "learning_rate": 6.076693150205601e-05, "loss": 0.0398, "step": 73500 }, { "epoch": 1.468966068502458, "grad_norm": 0.12309998273849487, "learning_rate": 6.07568395316131e-05, "loss": 0.0339, "step": 73510 }, { "epoch": 1.4691659006434594, "grad_norm": 0.11063095927238464, "learning_rate": 6.0746747101657744e-05, "loss": 0.0311, "step": 73520 }, { "epoch": 1.469365732784461, "grad_norm": 0.10352417826652527, "learning_rate": 6.0736654212621104e-05, "loss": 0.0335, "step": 73530 }, { "epoch": 1.4695655649254626, "grad_norm": 0.12067357450723648, "learning_rate": 6.0726560864934324e-05, "loss": 0.0249, "step": 73540 }, { "epoch": 1.4697653970664641, "grad_norm": 0.12063389271497726, "learning_rate": 6.071646705902856e-05, "loss": 0.0344, "step": 73550 }, { "epoch": 1.4699652292074656, "grad_norm": 0.11865145713090897, "learning_rate": 6.070637279533502e-05, "loss": 0.0309, "step": 73560 }, { "epoch": 1.4701650613484674, "grad_norm": 0.10718624293804169, "learning_rate": 6.069627807428489e-05, "loss": 0.031, "step": 73570 }, { "epoch": 1.4703648934894689, "grad_norm": 0.12219756841659546, "learning_rate": 6.068618289630943e-05, "loss": 0.0298, "step": 73580 }, { "epoch": 1.4705647256304704, "grad_norm": 0.1286827027797699, "learning_rate": 6.0676087261839844e-05, "loss": 0.0344, "step": 73590 }, { "epoch": 1.470764557771472, "grad_norm": 0.14045871794223785, "learning_rate": 6.066599117130745e-05, "loss": 0.0367, "step": 73600 }, { "epoch": 1.4709643899124736, "grad_norm": 0.11308804899454117, "learning_rate": 6.0655894625143485e-05, "loss": 0.0372, "step": 73610 }, { "epoch": 1.471164222053475, "grad_norm": 0.15700457990169525, "learning_rate": 6.0645797623779274e-05, "loss": 0.035, "step": 73620 }, { "epoch": 1.4713640541944766, "grad_norm": 0.12259246408939362, "learning_rate": 6.0635700167646156e-05, "loss": 0.0338, "step": 73630 }, { "epoch": 1.471563886335478, "grad_norm": 0.18062029778957367, "learning_rate": 6.062560225717546e-05, "loss": 0.0297, "step": 73640 }, { "epoch": 1.4717637184764798, "grad_norm": 0.14129993319511414, "learning_rate": 6.0615503892798575e-05, "loss": 0.0283, "step": 73650 }, { "epoch": 1.4719635506174813, "grad_norm": 0.11974095553159714, "learning_rate": 6.060540507494685e-05, "loss": 0.029, "step": 73660 }, { "epoch": 1.4721633827584828, "grad_norm": 0.1484316885471344, "learning_rate": 6.05953058040517e-05, "loss": 0.0297, "step": 73670 }, { "epoch": 1.4723632148994845, "grad_norm": 0.1297162026166916, "learning_rate": 6.058520608054455e-05, "loss": 0.0355, "step": 73680 }, { "epoch": 1.472563047040486, "grad_norm": 0.10127046704292297, "learning_rate": 6.0575105904856847e-05, "loss": 0.0334, "step": 73690 }, { "epoch": 1.4727628791814875, "grad_norm": 0.1835208684206009, "learning_rate": 6.056500527742004e-05, "loss": 0.04, "step": 73700 }, { "epoch": 1.4729627113224892, "grad_norm": 0.13290420174598694, "learning_rate": 6.055490419866563e-05, "loss": 0.0301, "step": 73710 }, { "epoch": 1.4731625434634907, "grad_norm": 0.15637251734733582, "learning_rate": 6.054480266902508e-05, "loss": 0.034, "step": 73720 }, { "epoch": 1.4733623756044922, "grad_norm": 0.14282412827014923, "learning_rate": 6.053470068892996e-05, "loss": 0.0226, "step": 73730 }, { "epoch": 1.4735622077454937, "grad_norm": 0.144269198179245, "learning_rate": 6.052459825881176e-05, "loss": 0.0269, "step": 73740 }, { "epoch": 1.4737620398864952, "grad_norm": 0.1384468972682953, "learning_rate": 6.051449537910207e-05, "loss": 0.032, "step": 73750 }, { "epoch": 1.473961872027497, "grad_norm": 0.11228329688310623, "learning_rate": 6.0504392050232436e-05, "loss": 0.0306, "step": 73760 }, { "epoch": 1.4741617041684985, "grad_norm": 0.18467991054058075, "learning_rate": 6.049428827263448e-05, "loss": 0.0244, "step": 73770 }, { "epoch": 1.4743615363095, "grad_norm": 0.11275158077478409, "learning_rate": 6.048418404673981e-05, "loss": 0.0419, "step": 73780 }, { "epoch": 1.4745613684505017, "grad_norm": 0.09691493958234787, "learning_rate": 6.047407937298004e-05, "loss": 0.0232, "step": 73790 }, { "epoch": 1.4747612005915032, "grad_norm": 0.1622837632894516, "learning_rate": 6.046397425178687e-05, "loss": 0.0314, "step": 73800 }, { "epoch": 1.4749610327325047, "grad_norm": 0.13889847695827484, "learning_rate": 6.045386868359193e-05, "loss": 0.0317, "step": 73810 }, { "epoch": 1.4751608648735062, "grad_norm": 0.14662142097949982, "learning_rate": 6.044376266882692e-05, "loss": 0.0291, "step": 73820 }, { "epoch": 1.475360697014508, "grad_norm": 0.1510991007089615, "learning_rate": 6.043365620792356e-05, "loss": 0.0379, "step": 73830 }, { "epoch": 1.4755605291555094, "grad_norm": 0.10576393455266953, "learning_rate": 6.042354930131356e-05, "loss": 0.0342, "step": 73840 }, { "epoch": 1.475760361296511, "grad_norm": 0.10690503567457199, "learning_rate": 6.0413441949428706e-05, "loss": 0.0299, "step": 73850 }, { "epoch": 1.4759601934375124, "grad_norm": 0.10433241724967957, "learning_rate": 6.040333415270072e-05, "loss": 0.0335, "step": 73860 }, { "epoch": 1.4761600255785141, "grad_norm": 0.12141530215740204, "learning_rate": 6.0393225911561426e-05, "loss": 0.0307, "step": 73870 }, { "epoch": 1.4763598577195156, "grad_norm": 0.136936217546463, "learning_rate": 6.038311722644261e-05, "loss": 0.0213, "step": 73880 }, { "epoch": 1.476559689860517, "grad_norm": 0.1338803917169571, "learning_rate": 6.0373008097776084e-05, "loss": 0.0317, "step": 73890 }, { "epoch": 1.4767595220015188, "grad_norm": 0.09977176785469055, "learning_rate": 6.036289852599373e-05, "loss": 0.0269, "step": 73900 }, { "epoch": 1.4769593541425203, "grad_norm": 0.1049320176243782, "learning_rate": 6.0352788511527394e-05, "loss": 0.0297, "step": 73910 }, { "epoch": 1.4771591862835218, "grad_norm": 0.11277533322572708, "learning_rate": 6.0342678054808934e-05, "loss": 0.0305, "step": 73920 }, { "epoch": 1.4773590184245233, "grad_norm": 0.14362427592277527, "learning_rate": 6.033256715627027e-05, "loss": 0.0251, "step": 73930 }, { "epoch": 1.4775588505655248, "grad_norm": 0.106343112885952, "learning_rate": 6.032245581634332e-05, "loss": 0.0355, "step": 73940 }, { "epoch": 1.4777586827065265, "grad_norm": 0.13303448259830475, "learning_rate": 6.0312344035460013e-05, "loss": 0.0279, "step": 73950 }, { "epoch": 1.477958514847528, "grad_norm": 0.0915306806564331, "learning_rate": 6.0302231814052335e-05, "loss": 0.0317, "step": 73960 }, { "epoch": 1.4781583469885295, "grad_norm": 0.13718366622924805, "learning_rate": 6.029211915255221e-05, "loss": 0.0338, "step": 73970 }, { "epoch": 1.4783581791295313, "grad_norm": 0.13224360346794128, "learning_rate": 6.0282006051391684e-05, "loss": 0.0292, "step": 73980 }, { "epoch": 1.4785580112705328, "grad_norm": 0.1347973793745041, "learning_rate": 6.027189251100274e-05, "loss": 0.0359, "step": 73990 }, { "epoch": 1.4787578434115343, "grad_norm": 0.11806128174066544, "learning_rate": 6.026177853181743e-05, "loss": 0.0259, "step": 74000 }, { "epoch": 1.478957675552536, "grad_norm": 0.14905671775341034, "learning_rate": 6.025166411426777e-05, "loss": 0.0362, "step": 74010 }, { "epoch": 1.4791575076935375, "grad_norm": 0.10811510682106018, "learning_rate": 6.024154925878587e-05, "loss": 0.0352, "step": 74020 }, { "epoch": 1.479357339834539, "grad_norm": 0.14270071685314178, "learning_rate": 6.0231433965803796e-05, "loss": 0.0321, "step": 74030 }, { "epoch": 1.4795571719755405, "grad_norm": 0.14582201838493347, "learning_rate": 6.0221318235753645e-05, "loss": 0.0339, "step": 74040 }, { "epoch": 1.479757004116542, "grad_norm": 0.11860936135053635, "learning_rate": 6.021120206906757e-05, "loss": 0.034, "step": 74050 }, { "epoch": 1.4799568362575437, "grad_norm": 0.1343764066696167, "learning_rate": 6.020108546617771e-05, "loss": 0.0304, "step": 74060 }, { "epoch": 1.4801566683985452, "grad_norm": 0.10121891647577286, "learning_rate": 6.01909684275162e-05, "loss": 0.0338, "step": 74070 }, { "epoch": 1.4803565005395467, "grad_norm": 0.11505205184221268, "learning_rate": 6.018085095351524e-05, "loss": 0.0281, "step": 74080 }, { "epoch": 1.4805563326805484, "grad_norm": 0.120442233979702, "learning_rate": 6.017073304460704e-05, "loss": 0.0326, "step": 74090 }, { "epoch": 1.48075616482155, "grad_norm": 0.14696131646633148, "learning_rate": 6.0160614701223814e-05, "loss": 0.0248, "step": 74100 }, { "epoch": 1.4809559969625514, "grad_norm": 0.1580084264278412, "learning_rate": 6.015049592379779e-05, "loss": 0.0293, "step": 74110 }, { "epoch": 1.4811558291035531, "grad_norm": 0.17828816175460815, "learning_rate": 6.014037671276123e-05, "loss": 0.0294, "step": 74120 }, { "epoch": 1.4813556612445546, "grad_norm": 0.1392558366060257, "learning_rate": 6.013025706854639e-05, "loss": 0.0307, "step": 74130 }, { "epoch": 1.4815554933855561, "grad_norm": 0.09969157725572586, "learning_rate": 6.012013699158558e-05, "loss": 0.0346, "step": 74140 }, { "epoch": 1.4817553255265576, "grad_norm": 0.15038813650608063, "learning_rate": 6.011001648231113e-05, "loss": 0.0327, "step": 74150 }, { "epoch": 1.4819551576675591, "grad_norm": 0.19022125005722046, "learning_rate": 6.009989554115534e-05, "loss": 0.041, "step": 74160 }, { "epoch": 1.4821549898085609, "grad_norm": 0.159031480550766, "learning_rate": 6.008977416855055e-05, "loss": 0.0283, "step": 74170 }, { "epoch": 1.4823548219495624, "grad_norm": 0.13521277904510498, "learning_rate": 6.0079652364929165e-05, "loss": 0.0367, "step": 74180 }, { "epoch": 1.4825546540905639, "grad_norm": 0.12573295831680298, "learning_rate": 6.0069530130723516e-05, "loss": 0.0302, "step": 74190 }, { "epoch": 1.4827544862315656, "grad_norm": 0.13587674498558044, "learning_rate": 6.005940746636607e-05, "loss": 0.034, "step": 74200 }, { "epoch": 1.482954318372567, "grad_norm": 0.1563725769519806, "learning_rate": 6.0049284372289206e-05, "loss": 0.0304, "step": 74210 }, { "epoch": 1.4831541505135686, "grad_norm": 0.1330161690711975, "learning_rate": 6.003916084892538e-05, "loss": 0.0334, "step": 74220 }, { "epoch": 1.48335398265457, "grad_norm": 0.1589941680431366, "learning_rate": 6.002903689670705e-05, "loss": 0.0284, "step": 74230 }, { "epoch": 1.4835538147955718, "grad_norm": 0.1525958627462387, "learning_rate": 6.001891251606666e-05, "loss": 0.028, "step": 74240 }, { "epoch": 1.4837536469365733, "grad_norm": 0.12344515323638916, "learning_rate": 6.000878770743675e-05, "loss": 0.0346, "step": 74250 }, { "epoch": 1.4839534790775748, "grad_norm": 0.13156816363334656, "learning_rate": 5.9998662471249824e-05, "loss": 0.0321, "step": 74260 }, { "epoch": 1.4841533112185763, "grad_norm": 0.1242574080824852, "learning_rate": 5.998853680793839e-05, "loss": 0.0361, "step": 74270 }, { "epoch": 1.484353143359578, "grad_norm": 0.1516328752040863, "learning_rate": 5.997841071793502e-05, "loss": 0.0367, "step": 74280 }, { "epoch": 1.4845529755005795, "grad_norm": 0.11945547163486481, "learning_rate": 5.996828420167227e-05, "loss": 0.0318, "step": 74290 }, { "epoch": 1.484752807641581, "grad_norm": 0.11482664197683334, "learning_rate": 5.995815725958276e-05, "loss": 0.0284, "step": 74300 }, { "epoch": 1.4849526397825827, "grad_norm": 0.11649198830127716, "learning_rate": 5.994802989209903e-05, "loss": 0.0272, "step": 74310 }, { "epoch": 1.4851524719235842, "grad_norm": 0.10048938542604446, "learning_rate": 5.993790209965375e-05, "loss": 0.023, "step": 74320 }, { "epoch": 1.4853523040645857, "grad_norm": 0.13401925563812256, "learning_rate": 5.992777388267955e-05, "loss": 0.0328, "step": 74330 }, { "epoch": 1.4855521362055872, "grad_norm": 0.08802513778209686, "learning_rate": 5.991764524160909e-05, "loss": 0.0261, "step": 74340 }, { "epoch": 1.4857519683465887, "grad_norm": 0.13391225039958954, "learning_rate": 5.9907516176875034e-05, "loss": 0.0383, "step": 74350 }, { "epoch": 1.4859518004875905, "grad_norm": 0.1238613948225975, "learning_rate": 5.9897386688910106e-05, "loss": 0.0337, "step": 74360 }, { "epoch": 1.486151632628592, "grad_norm": 0.13563044369220734, "learning_rate": 5.9887256778146974e-05, "loss": 0.0326, "step": 74370 }, { "epoch": 1.4863514647695935, "grad_norm": 0.16278736293315887, "learning_rate": 5.9877126445018426e-05, "loss": 0.0331, "step": 74380 }, { "epoch": 1.4865512969105952, "grad_norm": 0.15514500439167023, "learning_rate": 5.986699568995716e-05, "loss": 0.0291, "step": 74390 }, { "epoch": 1.4867511290515967, "grad_norm": 0.16239838302135468, "learning_rate": 5.9856864513395974e-05, "loss": 0.0353, "step": 74400 }, { "epoch": 1.4869509611925982, "grad_norm": 0.12458094954490662, "learning_rate": 5.9846732915767636e-05, "loss": 0.0399, "step": 74410 }, { "epoch": 1.4871507933336, "grad_norm": 0.15192440152168274, "learning_rate": 5.983660089750496e-05, "loss": 0.0261, "step": 74420 }, { "epoch": 1.4873506254746014, "grad_norm": 0.11990486085414886, "learning_rate": 5.982646845904078e-05, "loss": 0.0326, "step": 74430 }, { "epoch": 1.487550457615603, "grad_norm": 0.1121152713894844, "learning_rate": 5.981633560080789e-05, "loss": 0.0277, "step": 74440 }, { "epoch": 1.4877502897566044, "grad_norm": 0.11248280853033066, "learning_rate": 5.98062023232392e-05, "loss": 0.031, "step": 74450 }, { "epoch": 1.487950121897606, "grad_norm": 0.11609367281198502, "learning_rate": 5.979606862676755e-05, "loss": 0.0271, "step": 74460 }, { "epoch": 1.4881499540386076, "grad_norm": 0.10805945098400116, "learning_rate": 5.9785934511825844e-05, "loss": 0.0273, "step": 74470 }, { "epoch": 1.4883497861796091, "grad_norm": 0.10820624977350235, "learning_rate": 5.977579997884702e-05, "loss": 0.0285, "step": 74480 }, { "epoch": 1.4885496183206106, "grad_norm": 0.11292063444852829, "learning_rate": 5.976566502826393e-05, "loss": 0.0328, "step": 74490 }, { "epoch": 1.4887494504616123, "grad_norm": 0.12232398241758347, "learning_rate": 5.975552966050961e-05, "loss": 0.0285, "step": 74500 }, { "epoch": 1.4889492826026138, "grad_norm": 0.1007363498210907, "learning_rate": 5.9745393876016964e-05, "loss": 0.0299, "step": 74510 }, { "epoch": 1.4891491147436153, "grad_norm": 0.12561099231243134, "learning_rate": 5.9735257675219e-05, "loss": 0.028, "step": 74520 }, { "epoch": 1.489348946884617, "grad_norm": 0.09410183876752853, "learning_rate": 5.972512105854872e-05, "loss": 0.0246, "step": 74530 }, { "epoch": 1.4895487790256186, "grad_norm": 0.11683990806341171, "learning_rate": 5.971498402643911e-05, "loss": 0.0299, "step": 74540 }, { "epoch": 1.48974861116662, "grad_norm": 0.11098486930131912, "learning_rate": 5.970484657932325e-05, "loss": 0.0364, "step": 74550 }, { "epoch": 1.4899484433076216, "grad_norm": 0.08721994608640671, "learning_rate": 5.969470871763418e-05, "loss": 0.0267, "step": 74560 }, { "epoch": 1.490148275448623, "grad_norm": 0.11396309733390808, "learning_rate": 5.968457044180496e-05, "loss": 0.024, "step": 74570 }, { "epoch": 1.4903481075896248, "grad_norm": 0.11051981151103973, "learning_rate": 5.967443175226867e-05, "loss": 0.0281, "step": 74580 }, { "epoch": 1.4905479397306263, "grad_norm": 0.09569942206144333, "learning_rate": 5.966429264945842e-05, "loss": 0.0233, "step": 74590 }, { "epoch": 1.4907477718716278, "grad_norm": 0.10813800245523453, "learning_rate": 5.9654153133807354e-05, "loss": 0.0313, "step": 74600 }, { "epoch": 1.4909476040126295, "grad_norm": 0.14825540781021118, "learning_rate": 5.964401320574861e-05, "loss": 0.0254, "step": 74610 }, { "epoch": 1.491147436153631, "grad_norm": 0.11795242875814438, "learning_rate": 5.9633872865715325e-05, "loss": 0.0356, "step": 74620 }, { "epoch": 1.4913472682946325, "grad_norm": 0.14515064656734467, "learning_rate": 5.9623732114140696e-05, "loss": 0.026, "step": 74630 }, { "epoch": 1.491547100435634, "grad_norm": 0.11312663555145264, "learning_rate": 5.96135909514579e-05, "loss": 0.0261, "step": 74640 }, { "epoch": 1.4917469325766357, "grad_norm": 0.10606344789266586, "learning_rate": 5.960344937810017e-05, "loss": 0.0268, "step": 74650 }, { "epoch": 1.4919467647176372, "grad_norm": 0.1636648029088974, "learning_rate": 5.959330739450073e-05, "loss": 0.0259, "step": 74660 }, { "epoch": 1.4921465968586387, "grad_norm": 0.11525661498308182, "learning_rate": 5.958316500109281e-05, "loss": 0.0277, "step": 74670 }, { "epoch": 1.4923464289996402, "grad_norm": 0.11528882384300232, "learning_rate": 5.957302219830969e-05, "loss": 0.0222, "step": 74680 }, { "epoch": 1.492546261140642, "grad_norm": 0.11194194108247757, "learning_rate": 5.956287898658464e-05, "loss": 0.0347, "step": 74690 }, { "epoch": 1.4927460932816434, "grad_norm": 0.11596120893955231, "learning_rate": 5.9552735366350985e-05, "loss": 0.0333, "step": 74700 }, { "epoch": 1.492945925422645, "grad_norm": 0.1061655730009079, "learning_rate": 5.954259133804201e-05, "loss": 0.0336, "step": 74710 }, { "epoch": 1.4931457575636466, "grad_norm": 0.149813711643219, "learning_rate": 5.953244690209106e-05, "loss": 0.0287, "step": 74720 }, { "epoch": 1.4933455897046481, "grad_norm": 0.1440758854150772, "learning_rate": 5.95223020589315e-05, "loss": 0.0294, "step": 74730 }, { "epoch": 1.4935454218456496, "grad_norm": 0.09859760850667953, "learning_rate": 5.951215680899669e-05, "loss": 0.0334, "step": 74740 }, { "epoch": 1.4937452539866511, "grad_norm": 0.12680160999298096, "learning_rate": 5.9502011152720004e-05, "loss": 0.0412, "step": 74750 }, { "epoch": 1.4939450861276526, "grad_norm": 0.11889845877885818, "learning_rate": 5.949186509053488e-05, "loss": 0.0291, "step": 74760 }, { "epoch": 1.4941449182686544, "grad_norm": 0.1299707591533661, "learning_rate": 5.94817186228747e-05, "loss": 0.026, "step": 74770 }, { "epoch": 1.4943447504096559, "grad_norm": 0.11905336380004883, "learning_rate": 5.9471571750172916e-05, "loss": 0.0294, "step": 74780 }, { "epoch": 1.4945445825506574, "grad_norm": 0.12642724812030792, "learning_rate": 5.946142447286299e-05, "loss": 0.0382, "step": 74790 }, { "epoch": 1.494744414691659, "grad_norm": 0.1158878281712532, "learning_rate": 5.9451276791378396e-05, "loss": 0.0342, "step": 74800 }, { "epoch": 1.4949442468326606, "grad_norm": 0.1029149740934372, "learning_rate": 5.944112870615262e-05, "loss": 0.0204, "step": 74810 }, { "epoch": 1.495144078973662, "grad_norm": 0.15126727521419525, "learning_rate": 5.943098021761916e-05, "loss": 0.0342, "step": 74820 }, { "epoch": 1.4953439111146638, "grad_norm": 0.15377569198608398, "learning_rate": 5.942083132621157e-05, "loss": 0.0348, "step": 74830 }, { "epoch": 1.4955437432556653, "grad_norm": 0.14769220352172852, "learning_rate": 5.941068203236334e-05, "loss": 0.034, "step": 74840 }, { "epoch": 1.4957435753966668, "grad_norm": 0.12450861930847168, "learning_rate": 5.94005323365081e-05, "loss": 0.0283, "step": 74850 }, { "epoch": 1.4959434075376683, "grad_norm": 0.126613587141037, "learning_rate": 5.9390382239079356e-05, "loss": 0.0372, "step": 74860 }, { "epoch": 1.4961432396786698, "grad_norm": 0.12484042346477509, "learning_rate": 5.938023174051075e-05, "loss": 0.0272, "step": 74870 }, { "epoch": 1.4963430718196715, "grad_norm": 0.12464787065982819, "learning_rate": 5.9370080841235875e-05, "loss": 0.0367, "step": 74880 }, { "epoch": 1.496542903960673, "grad_norm": 0.12777329981327057, "learning_rate": 5.935992954168835e-05, "loss": 0.0288, "step": 74890 }, { "epoch": 1.4967427361016745, "grad_norm": 0.12153763324022293, "learning_rate": 5.9349777842301844e-05, "loss": 0.0294, "step": 74900 }, { "epoch": 1.4969425682426762, "grad_norm": 0.1373182237148285, "learning_rate": 5.9339625743509995e-05, "loss": 0.0269, "step": 74910 }, { "epoch": 1.4971424003836777, "grad_norm": 0.1364590972661972, "learning_rate": 5.932947324574649e-05, "loss": 0.027, "step": 74920 }, { "epoch": 1.4973422325246792, "grad_norm": 0.11283166706562042, "learning_rate": 5.931932034944504e-05, "loss": 0.0366, "step": 74930 }, { "epoch": 1.497542064665681, "grad_norm": 0.13744714856147766, "learning_rate": 5.930916705503933e-05, "loss": 0.0366, "step": 74940 }, { "epoch": 1.4977418968066825, "grad_norm": 0.13862909376621246, "learning_rate": 5.9299013362963116e-05, "loss": 0.0337, "step": 74950 }, { "epoch": 1.497941728947684, "grad_norm": 0.12385968863964081, "learning_rate": 5.928885927365013e-05, "loss": 0.0242, "step": 74960 }, { "epoch": 1.4981415610886855, "grad_norm": 0.12370108813047409, "learning_rate": 5.9278704787534144e-05, "loss": 0.0379, "step": 74970 }, { "epoch": 1.498341393229687, "grad_norm": 0.12337838113307953, "learning_rate": 5.926854990504892e-05, "loss": 0.0287, "step": 74980 }, { "epoch": 1.4985412253706887, "grad_norm": 0.12818856537342072, "learning_rate": 5.925839462662828e-05, "loss": 0.0357, "step": 74990 }, { "epoch": 1.4987410575116902, "grad_norm": 0.13689669966697693, "learning_rate": 5.924823895270604e-05, "loss": 0.0288, "step": 75000 }, { "epoch": 1.4989408896526917, "grad_norm": 0.11161930859088898, "learning_rate": 5.923808288371601e-05, "loss": 0.0319, "step": 75010 }, { "epoch": 1.4991407217936934, "grad_norm": 0.13290627300739288, "learning_rate": 5.922792642009204e-05, "loss": 0.0304, "step": 75020 }, { "epoch": 1.499340553934695, "grad_norm": 0.11144503951072693, "learning_rate": 5.9217769562268035e-05, "loss": 0.0304, "step": 75030 }, { "epoch": 1.4995403860756964, "grad_norm": 0.13928203284740448, "learning_rate": 5.920761231067782e-05, "loss": 0.0258, "step": 75040 }, { "epoch": 1.499740218216698, "grad_norm": 0.15306876599788666, "learning_rate": 5.919745466575534e-05, "loss": 0.0316, "step": 75050 }, { "epoch": 1.4999400503576994, "grad_norm": 0.11548259854316711, "learning_rate": 5.9187296627934476e-05, "loss": 0.0305, "step": 75060 }, { "epoch": 1.5001398824987011, "grad_norm": 0.15868759155273438, "learning_rate": 5.9177138197649186e-05, "loss": 0.0335, "step": 75070 }, { "epoch": 1.5003397146397026, "grad_norm": 0.16888564825057983, "learning_rate": 5.916697937533342e-05, "loss": 0.0295, "step": 75080 }, { "epoch": 1.5005395467807041, "grad_norm": 0.13597561419010162, "learning_rate": 5.9156820161421126e-05, "loss": 0.0349, "step": 75090 }, { "epoch": 1.5007393789217058, "grad_norm": 0.12335336953401566, "learning_rate": 5.91466605563463e-05, "loss": 0.0274, "step": 75100 }, { "epoch": 1.5009392110627073, "grad_norm": 0.19747306406497955, "learning_rate": 5.913650056054293e-05, "loss": 0.0295, "step": 75110 }, { "epoch": 1.5011390432037088, "grad_norm": 0.12882421910762787, "learning_rate": 5.912634017444505e-05, "loss": 0.0287, "step": 75120 }, { "epoch": 1.5013388753447106, "grad_norm": 0.13951727747917175, "learning_rate": 5.911617939848668e-05, "loss": 0.0287, "step": 75130 }, { "epoch": 1.501538707485712, "grad_norm": 0.11670302599668503, "learning_rate": 5.910601823310187e-05, "loss": 0.0345, "step": 75140 }, { "epoch": 1.5017385396267136, "grad_norm": 0.13092947006225586, "learning_rate": 5.9095856678724706e-05, "loss": 0.0231, "step": 75150 }, { "epoch": 1.5019383717677153, "grad_norm": 0.10444718599319458, "learning_rate": 5.908569473578923e-05, "loss": 0.0256, "step": 75160 }, { "epoch": 1.5021382039087166, "grad_norm": 0.14219585061073303, "learning_rate": 5.907553240472957e-05, "loss": 0.0339, "step": 75170 }, { "epoch": 1.5023380360497183, "grad_norm": 0.17175006866455078, "learning_rate": 5.9065369685979845e-05, "loss": 0.0336, "step": 75180 }, { "epoch": 1.5025378681907198, "grad_norm": 0.13809938728809357, "learning_rate": 5.9055206579974186e-05, "loss": 0.0274, "step": 75190 }, { "epoch": 1.5027377003317213, "grad_norm": 0.152793750166893, "learning_rate": 5.9045043087146726e-05, "loss": 0.0378, "step": 75200 }, { "epoch": 1.502937532472723, "grad_norm": 0.1394161432981491, "learning_rate": 5.903487920793165e-05, "loss": 0.0311, "step": 75210 }, { "epoch": 1.5031373646137245, "grad_norm": 0.11020372807979584, "learning_rate": 5.902471494276313e-05, "loss": 0.0291, "step": 75220 }, { "epoch": 1.503337196754726, "grad_norm": 0.11599457263946533, "learning_rate": 5.9014550292075363e-05, "loss": 0.0279, "step": 75230 }, { "epoch": 1.5035370288957277, "grad_norm": 0.11167824268341064, "learning_rate": 5.900438525630256e-05, "loss": 0.0262, "step": 75240 }, { "epoch": 1.503736861036729, "grad_norm": 0.14309875667095184, "learning_rate": 5.899421983587897e-05, "loss": 0.0316, "step": 75250 }, { "epoch": 1.5039366931777307, "grad_norm": 0.12009470164775848, "learning_rate": 5.898405403123884e-05, "loss": 0.0281, "step": 75260 }, { "epoch": 1.5041365253187324, "grad_norm": 0.10796387493610382, "learning_rate": 5.897388784281641e-05, "loss": 0.0307, "step": 75270 }, { "epoch": 1.5043363574597337, "grad_norm": 0.13982854783535004, "learning_rate": 5.896372127104599e-05, "loss": 0.0388, "step": 75280 }, { "epoch": 1.5045361896007354, "grad_norm": 0.17004144191741943, "learning_rate": 5.895355431636185e-05, "loss": 0.0361, "step": 75290 }, { "epoch": 1.504736021741737, "grad_norm": 0.11399630457162857, "learning_rate": 5.8943386979198334e-05, "loss": 0.026, "step": 75300 }, { "epoch": 1.5049358538827384, "grad_norm": 0.14683206379413605, "learning_rate": 5.893321925998975e-05, "loss": 0.0266, "step": 75310 }, { "epoch": 1.5051356860237401, "grad_norm": 0.13247129321098328, "learning_rate": 5.892305115917045e-05, "loss": 0.0336, "step": 75320 }, { "epoch": 1.5053355181647416, "grad_norm": 0.1281135380268097, "learning_rate": 5.8912882677174795e-05, "loss": 0.0306, "step": 75330 }, { "epoch": 1.5055353503057431, "grad_norm": 0.1626536250114441, "learning_rate": 5.890271381443716e-05, "loss": 0.0353, "step": 75340 }, { "epoch": 1.5057351824467449, "grad_norm": 0.1029878482222557, "learning_rate": 5.889254457139196e-05, "loss": 0.0382, "step": 75350 }, { "epoch": 1.5059350145877461, "grad_norm": 0.1499248445034027, "learning_rate": 5.888237494847357e-05, "loss": 0.027, "step": 75360 }, { "epoch": 1.5061348467287479, "grad_norm": 0.14296938478946686, "learning_rate": 5.887220494611646e-05, "loss": 0.0346, "step": 75370 }, { "epoch": 1.5063346788697494, "grad_norm": 0.13619501888751984, "learning_rate": 5.886203456475504e-05, "loss": 0.0234, "step": 75380 }, { "epoch": 1.5065345110107509, "grad_norm": 0.12099060416221619, "learning_rate": 5.8851863804823795e-05, "loss": 0.0323, "step": 75390 }, { "epoch": 1.5067343431517526, "grad_norm": 0.11215914040803909, "learning_rate": 5.884169266675718e-05, "loss": 0.0395, "step": 75400 }, { "epoch": 1.506934175292754, "grad_norm": 0.12193727493286133, "learning_rate": 5.88315211509897e-05, "loss": 0.0265, "step": 75410 }, { "epoch": 1.5071340074337556, "grad_norm": 0.15086764097213745, "learning_rate": 5.8821349257955874e-05, "loss": 0.0356, "step": 75420 }, { "epoch": 1.5073338395747573, "grad_norm": 0.12803903222084045, "learning_rate": 5.88111769880902e-05, "loss": 0.0361, "step": 75430 }, { "epoch": 1.5075336717157588, "grad_norm": 0.11369660496711731, "learning_rate": 5.8801004341827226e-05, "loss": 0.0259, "step": 75440 }, { "epoch": 1.5077335038567603, "grad_norm": 0.14614583551883698, "learning_rate": 5.879083131960153e-05, "loss": 0.0305, "step": 75450 }, { "epoch": 1.507933335997762, "grad_norm": 0.13493114709854126, "learning_rate": 5.878065792184765e-05, "loss": 0.0302, "step": 75460 }, { "epoch": 1.5081331681387633, "grad_norm": 0.16495755314826965, "learning_rate": 5.87704841490002e-05, "loss": 0.0291, "step": 75470 }, { "epoch": 1.508333000279765, "grad_norm": 0.111658476293087, "learning_rate": 5.876031000149379e-05, "loss": 0.0292, "step": 75480 }, { "epoch": 1.5085328324207665, "grad_norm": 0.1484554558992386, "learning_rate": 5.875013547976301e-05, "loss": 0.0267, "step": 75490 }, { "epoch": 1.508732664561768, "grad_norm": 0.11650984734296799, "learning_rate": 5.873996058424255e-05, "loss": 0.0307, "step": 75500 }, { "epoch": 1.5089324967027697, "grad_norm": 0.12943387031555176, "learning_rate": 5.8729785315367e-05, "loss": 0.0307, "step": 75510 }, { "epoch": 1.5091323288437712, "grad_norm": 0.14016661047935486, "learning_rate": 5.8719609673571066e-05, "loss": 0.0299, "step": 75520 }, { "epoch": 1.5093321609847727, "grad_norm": 0.12025056034326553, "learning_rate": 5.870943365928944e-05, "loss": 0.0407, "step": 75530 }, { "epoch": 1.5095319931257745, "grad_norm": 0.14470045268535614, "learning_rate": 5.8699257272956785e-05, "loss": 0.026, "step": 75540 }, { "epoch": 1.509731825266776, "grad_norm": 0.1095915138721466, "learning_rate": 5.8689080515007855e-05, "loss": 0.0245, "step": 75550 }, { "epoch": 1.5099316574077775, "grad_norm": 0.11174586415290833, "learning_rate": 5.867890338587736e-05, "loss": 0.0303, "step": 75560 }, { "epoch": 1.5101314895487792, "grad_norm": 0.12856905162334442, "learning_rate": 5.8668725886000066e-05, "loss": 0.0326, "step": 75570 }, { "epoch": 1.5103313216897805, "grad_norm": 0.13835328817367554, "learning_rate": 5.865854801581071e-05, "loss": 0.0255, "step": 75580 }, { "epoch": 1.5105311538307822, "grad_norm": 0.14793729782104492, "learning_rate": 5.864836977574411e-05, "loss": 0.027, "step": 75590 }, { "epoch": 1.5107309859717837, "grad_norm": 0.16212430596351624, "learning_rate": 5.863819116623503e-05, "loss": 0.0266, "step": 75600 }, { "epoch": 1.5109308181127852, "grad_norm": 0.18715347349643707, "learning_rate": 5.8628012187718284e-05, "loss": 0.0259, "step": 75610 }, { "epoch": 1.511130650253787, "grad_norm": 0.1487760841846466, "learning_rate": 5.861783284062873e-05, "loss": 0.0306, "step": 75620 }, { "epoch": 1.5113304823947884, "grad_norm": 0.11925332993268967, "learning_rate": 5.8607653125401164e-05, "loss": 0.0262, "step": 75630 }, { "epoch": 1.51153031453579, "grad_norm": 0.11759424209594727, "learning_rate": 5.8597473042470485e-05, "loss": 0.0401, "step": 75640 }, { "epoch": 1.5117301466767916, "grad_norm": 0.15139970183372498, "learning_rate": 5.858729259227154e-05, "loss": 0.0286, "step": 75650 }, { "epoch": 1.511929978817793, "grad_norm": 0.10490910708904266, "learning_rate": 5.8577111775239236e-05, "loss": 0.0384, "step": 75660 }, { "epoch": 1.5121298109587946, "grad_norm": 0.18981921672821045, "learning_rate": 5.8566930591808475e-05, "loss": 0.0303, "step": 75670 }, { "epoch": 1.5123296430997961, "grad_norm": 0.17371471226215363, "learning_rate": 5.855674904241418e-05, "loss": 0.0369, "step": 75680 }, { "epoch": 1.5125294752407976, "grad_norm": 0.13595466315746307, "learning_rate": 5.854656712749128e-05, "loss": 0.0275, "step": 75690 }, { "epoch": 1.5127293073817993, "grad_norm": 0.14073623716831207, "learning_rate": 5.853638484747473e-05, "loss": 0.0301, "step": 75700 }, { "epoch": 1.5129291395228008, "grad_norm": 0.11591969430446625, "learning_rate": 5.8526202202799516e-05, "loss": 0.0298, "step": 75710 }, { "epoch": 1.5131289716638023, "grad_norm": 0.1469113826751709, "learning_rate": 5.851601919390059e-05, "loss": 0.0278, "step": 75720 }, { "epoch": 1.513328803804804, "grad_norm": 0.13457055389881134, "learning_rate": 5.8505835821212985e-05, "loss": 0.028, "step": 75730 }, { "epoch": 1.5135286359458056, "grad_norm": 0.1440034806728363, "learning_rate": 5.849565208517168e-05, "loss": 0.0342, "step": 75740 }, { "epoch": 1.513728468086807, "grad_norm": 0.1241665631532669, "learning_rate": 5.8485467986211754e-05, "loss": 0.0245, "step": 75750 }, { "epoch": 1.5139283002278088, "grad_norm": 0.11480426043272018, "learning_rate": 5.847528352476819e-05, "loss": 0.0224, "step": 75760 }, { "epoch": 1.51412813236881, "grad_norm": 0.08676614612340927, "learning_rate": 5.84650987012761e-05, "loss": 0.0252, "step": 75770 }, { "epoch": 1.5143279645098118, "grad_norm": 0.11070332676172256, "learning_rate": 5.845491351617055e-05, "loss": 0.0282, "step": 75780 }, { "epoch": 1.5145277966508133, "grad_norm": 0.1075819805264473, "learning_rate": 5.844472796988661e-05, "loss": 0.0302, "step": 75790 }, { "epoch": 1.5147276287918148, "grad_norm": 0.14931967854499817, "learning_rate": 5.843454206285942e-05, "loss": 0.0367, "step": 75800 }, { "epoch": 1.5149274609328165, "grad_norm": 0.10356732457876205, "learning_rate": 5.842435579552408e-05, "loss": 0.0265, "step": 75810 }, { "epoch": 1.515127293073818, "grad_norm": 0.13343890011310577, "learning_rate": 5.8414169168315735e-05, "loss": 0.0264, "step": 75820 }, { "epoch": 1.5153271252148195, "grad_norm": 0.12214706838130951, "learning_rate": 5.840398218166955e-05, "loss": 0.0359, "step": 75830 }, { "epoch": 1.5155269573558212, "grad_norm": 0.1504773199558258, "learning_rate": 5.839379483602068e-05, "loss": 0.0297, "step": 75840 }, { "epoch": 1.5157267894968227, "grad_norm": 0.12326496839523315, "learning_rate": 5.838360713180431e-05, "loss": 0.0392, "step": 75850 }, { "epoch": 1.5159266216378242, "grad_norm": 0.119497150182724, "learning_rate": 5.8373419069455636e-05, "loss": 0.0301, "step": 75860 }, { "epoch": 1.516126453778826, "grad_norm": 0.12958264350891113, "learning_rate": 5.8363230649409894e-05, "loss": 0.0291, "step": 75870 }, { "epoch": 1.5163262859198272, "grad_norm": 0.15414847433567047, "learning_rate": 5.835304187210231e-05, "loss": 0.0285, "step": 75880 }, { "epoch": 1.516526118060829, "grad_norm": 0.12156045436859131, "learning_rate": 5.834285273796811e-05, "loss": 0.033, "step": 75890 }, { "epoch": 1.5167259502018304, "grad_norm": 0.1226297989487648, "learning_rate": 5.833266324744257e-05, "loss": 0.0234, "step": 75900 }, { "epoch": 1.516925782342832, "grad_norm": 0.11734097450971603, "learning_rate": 5.832247340096098e-05, "loss": 0.0414, "step": 75910 }, { "epoch": 1.5171256144838337, "grad_norm": 0.15689992904663086, "learning_rate": 5.831228319895858e-05, "loss": 0.0296, "step": 75920 }, { "epoch": 1.5173254466248352, "grad_norm": 0.2186635434627533, "learning_rate": 5.8302092641870756e-05, "loss": 0.0306, "step": 75930 }, { "epoch": 1.5175252787658366, "grad_norm": 0.12966305017471313, "learning_rate": 5.829190173013275e-05, "loss": 0.0355, "step": 75940 }, { "epoch": 1.5177251109068384, "grad_norm": 0.13759535551071167, "learning_rate": 5.828171046417995e-05, "loss": 0.0238, "step": 75950 }, { "epoch": 1.5179249430478399, "grad_norm": 0.11313489824533463, "learning_rate": 5.827151884444768e-05, "loss": 0.038, "step": 75960 }, { "epoch": 1.5181247751888414, "grad_norm": 0.11620961874723434, "learning_rate": 5.8261326871371336e-05, "loss": 0.0244, "step": 75970 }, { "epoch": 1.518324607329843, "grad_norm": 0.15898926556110382, "learning_rate": 5.8251134545386275e-05, "loss": 0.0304, "step": 75980 }, { "epoch": 1.5185244394708444, "grad_norm": 0.10774866491556168, "learning_rate": 5.8240941866927914e-05, "loss": 0.024, "step": 75990 }, { "epoch": 1.518724271611846, "grad_norm": 0.1489928513765335, "learning_rate": 5.823074883643165e-05, "loss": 0.0332, "step": 76000 }, { "epoch": 1.5189241037528476, "grad_norm": 0.12644103169441223, "learning_rate": 5.82205554543329e-05, "loss": 0.0288, "step": 76010 }, { "epoch": 1.519123935893849, "grad_norm": 0.11531005799770355, "learning_rate": 5.8210361721067155e-05, "loss": 0.0333, "step": 76020 }, { "epoch": 1.5193237680348508, "grad_norm": 0.11945836246013641, "learning_rate": 5.820016763706981e-05, "loss": 0.0293, "step": 76030 }, { "epoch": 1.5195236001758523, "grad_norm": 0.13881386816501617, "learning_rate": 5.818997320277639e-05, "loss": 0.0259, "step": 76040 }, { "epoch": 1.5197234323168538, "grad_norm": 0.11571250855922699, "learning_rate": 5.8179778418622346e-05, "loss": 0.0256, "step": 76050 }, { "epoch": 1.5199232644578555, "grad_norm": 0.1160099059343338, "learning_rate": 5.816958328504321e-05, "loss": 0.0292, "step": 76060 }, { "epoch": 1.5201230965988568, "grad_norm": 0.19275113940238953, "learning_rate": 5.8159387802474476e-05, "loss": 0.0269, "step": 76070 }, { "epoch": 1.5203229287398585, "grad_norm": 0.11693526804447174, "learning_rate": 5.814919197135168e-05, "loss": 0.0353, "step": 76080 }, { "epoch": 1.52052276088086, "grad_norm": 0.13691996037960052, "learning_rate": 5.8138995792110375e-05, "loss": 0.0289, "step": 76090 }, { "epoch": 1.5207225930218615, "grad_norm": 0.11892274022102356, "learning_rate": 5.8128799265186116e-05, "loss": 0.0337, "step": 76100 }, { "epoch": 1.5209224251628632, "grad_norm": 0.130788192152977, "learning_rate": 5.8118602391014486e-05, "loss": 0.04, "step": 76110 }, { "epoch": 1.5211222573038647, "grad_norm": 0.12389378249645233, "learning_rate": 5.810840517003109e-05, "loss": 0.029, "step": 76120 }, { "epoch": 1.5213220894448662, "grad_norm": 0.14579561352729797, "learning_rate": 5.8098207602671516e-05, "loss": 0.0365, "step": 76130 }, { "epoch": 1.521521921585868, "grad_norm": 0.1303713470697403, "learning_rate": 5.808800968937138e-05, "loss": 0.0305, "step": 76140 }, { "epoch": 1.5217217537268695, "grad_norm": 0.11495066434144974, "learning_rate": 5.807781143056634e-05, "loss": 0.0351, "step": 76150 }, { "epoch": 1.521921585867871, "grad_norm": 0.11999022215604782, "learning_rate": 5.806761282669203e-05, "loss": 0.0377, "step": 76160 }, { "epoch": 1.5221214180088727, "grad_norm": 0.1031784936785698, "learning_rate": 5.805741387818412e-05, "loss": 0.0259, "step": 76170 }, { "epoch": 1.522321250149874, "grad_norm": 0.10490730404853821, "learning_rate": 5.804721458547828e-05, "loss": 0.0296, "step": 76180 }, { "epoch": 1.5225210822908757, "grad_norm": 0.11632727086544037, "learning_rate": 5.803701494901024e-05, "loss": 0.0366, "step": 76190 }, { "epoch": 1.5227209144318772, "grad_norm": 0.11720330268144608, "learning_rate": 5.802681496921567e-05, "loss": 0.0248, "step": 76200 }, { "epoch": 1.5229207465728787, "grad_norm": 0.12297004461288452, "learning_rate": 5.801661464653031e-05, "loss": 0.0347, "step": 76210 }, { "epoch": 1.5231205787138804, "grad_norm": 0.11020147055387497, "learning_rate": 5.8006413981389906e-05, "loss": 0.0407, "step": 76220 }, { "epoch": 1.523320410854882, "grad_norm": 0.11553619801998138, "learning_rate": 5.79962129742302e-05, "loss": 0.0317, "step": 76230 }, { "epoch": 1.5235202429958834, "grad_norm": 0.15714339911937714, "learning_rate": 5.798601162548697e-05, "loss": 0.0302, "step": 76240 }, { "epoch": 1.5237200751368851, "grad_norm": 0.16720248758792877, "learning_rate": 5.7975809935596e-05, "loss": 0.0309, "step": 76250 }, { "epoch": 1.5239199072778866, "grad_norm": 0.10311722755432129, "learning_rate": 5.796560790499306e-05, "loss": 0.0333, "step": 76260 }, { "epoch": 1.5241197394188881, "grad_norm": 0.12736541032791138, "learning_rate": 5.7955405534114004e-05, "loss": 0.0325, "step": 76270 }, { "epoch": 1.5243195715598898, "grad_norm": 0.14918135106563568, "learning_rate": 5.794520282339464e-05, "loss": 0.0256, "step": 76280 }, { "epoch": 1.5245194037008911, "grad_norm": 0.10378669202327728, "learning_rate": 5.7934999773270795e-05, "loss": 0.0279, "step": 76290 }, { "epoch": 1.5247192358418928, "grad_norm": 0.16007474064826965, "learning_rate": 5.792479638417834e-05, "loss": 0.0375, "step": 76300 }, { "epoch": 1.5249190679828943, "grad_norm": 0.1436692476272583, "learning_rate": 5.7914592656553134e-05, "loss": 0.0275, "step": 76310 }, { "epoch": 1.5251189001238958, "grad_norm": 0.10201804339885712, "learning_rate": 5.790438859083107e-05, "loss": 0.0324, "step": 76320 }, { "epoch": 1.5253187322648976, "grad_norm": 0.0908941999077797, "learning_rate": 5.789418418744806e-05, "loss": 0.0233, "step": 76330 }, { "epoch": 1.525518564405899, "grad_norm": 0.10862987488508224, "learning_rate": 5.7883979446840006e-05, "loss": 0.0348, "step": 76340 }, { "epoch": 1.5257183965469006, "grad_norm": 0.10188757628202438, "learning_rate": 5.787377436944282e-05, "loss": 0.023, "step": 76350 }, { "epoch": 1.5259182286879023, "grad_norm": 0.13234391808509827, "learning_rate": 5.786356895569246e-05, "loss": 0.034, "step": 76360 }, { "epoch": 1.5261180608289036, "grad_norm": 0.1362611949443817, "learning_rate": 5.785336320602488e-05, "loss": 0.038, "step": 76370 }, { "epoch": 1.5263178929699053, "grad_norm": 0.16549062728881836, "learning_rate": 5.7843157120876045e-05, "loss": 0.0276, "step": 76380 }, { "epoch": 1.526517725110907, "grad_norm": 0.14268949627876282, "learning_rate": 5.7832950700681954e-05, "loss": 0.0462, "step": 76390 }, { "epoch": 1.5267175572519083, "grad_norm": 0.14721369743347168, "learning_rate": 5.782274394587861e-05, "loss": 0.0296, "step": 76400 }, { "epoch": 1.52691738939291, "grad_norm": 0.18080563843250275, "learning_rate": 5.781253685690199e-05, "loss": 0.0404, "step": 76410 }, { "epoch": 1.5271172215339115, "grad_norm": 0.15544649958610535, "learning_rate": 5.7802329434188174e-05, "loss": 0.0395, "step": 76420 }, { "epoch": 1.527317053674913, "grad_norm": 0.15341104567050934, "learning_rate": 5.7792121678173164e-05, "loss": 0.0395, "step": 76430 }, { "epoch": 1.5275168858159147, "grad_norm": 0.1159035712480545, "learning_rate": 5.778191358929304e-05, "loss": 0.027, "step": 76440 }, { "epoch": 1.5277167179569162, "grad_norm": 0.1392306089401245, "learning_rate": 5.777170516798386e-05, "loss": 0.0344, "step": 76450 }, { "epoch": 1.5279165500979177, "grad_norm": 0.13315653800964355, "learning_rate": 5.776149641468169e-05, "loss": 0.0329, "step": 76460 }, { "epoch": 1.5281163822389194, "grad_norm": 0.14773832261562347, "learning_rate": 5.775128732982268e-05, "loss": 0.0272, "step": 76470 }, { "epoch": 1.5283162143799207, "grad_norm": 0.12417709827423096, "learning_rate": 5.77410779138429e-05, "loss": 0.0305, "step": 76480 }, { "epoch": 1.5285160465209224, "grad_norm": 0.12787088751792908, "learning_rate": 5.77308681671785e-05, "loss": 0.0297, "step": 76490 }, { "epoch": 1.528715878661924, "grad_norm": 0.13434426486492157, "learning_rate": 5.77206580902656e-05, "loss": 0.037, "step": 76500 }, { "epoch": 1.5289157108029254, "grad_norm": 0.17333167791366577, "learning_rate": 5.771044768354038e-05, "loss": 0.0295, "step": 76510 }, { "epoch": 1.5291155429439272, "grad_norm": 0.1190062016248703, "learning_rate": 5.7700236947439e-05, "loss": 0.0388, "step": 76520 }, { "epoch": 1.5293153750849287, "grad_norm": 0.13365551829338074, "learning_rate": 5.769002588239765e-05, "loss": 0.0241, "step": 76530 }, { "epoch": 1.5295152072259302, "grad_norm": 0.12469969689846039, "learning_rate": 5.767981448885251e-05, "loss": 0.0335, "step": 76540 }, { "epoch": 1.5297150393669319, "grad_norm": 0.13918042182922363, "learning_rate": 5.766960276723981e-05, "loss": 0.0267, "step": 76550 }, { "epoch": 1.5299148715079334, "grad_norm": 0.13713668286800385, "learning_rate": 5.765939071799576e-05, "loss": 0.0449, "step": 76560 }, { "epoch": 1.5301147036489349, "grad_norm": 0.11933567374944687, "learning_rate": 5.764917834155662e-05, "loss": 0.0248, "step": 76570 }, { "epoch": 1.5303145357899366, "grad_norm": 0.1406504064798355, "learning_rate": 5.763896563835862e-05, "loss": 0.0371, "step": 76580 }, { "epoch": 1.5305143679309379, "grad_norm": 0.10987545549869537, "learning_rate": 5.7628752608838055e-05, "loss": 0.0305, "step": 76590 }, { "epoch": 1.5307142000719396, "grad_norm": 0.12952455878257751, "learning_rate": 5.761853925343119e-05, "loss": 0.0277, "step": 76600 }, { "epoch": 1.530914032212941, "grad_norm": 0.1532459706068039, "learning_rate": 5.760832557257432e-05, "loss": 0.0327, "step": 76610 }, { "epoch": 1.5311138643539426, "grad_norm": 0.14341212809085846, "learning_rate": 5.759811156670377e-05, "loss": 0.043, "step": 76620 }, { "epoch": 1.5313136964949443, "grad_norm": 0.13822902739048004, "learning_rate": 5.7587897236255836e-05, "loss": 0.0229, "step": 76630 }, { "epoch": 1.5315135286359458, "grad_norm": 0.1449228823184967, "learning_rate": 5.757768258166688e-05, "loss": 0.0263, "step": 76640 }, { "epoch": 1.5317133607769473, "grad_norm": 0.11181005835533142, "learning_rate": 5.756746760337325e-05, "loss": 0.0251, "step": 76650 }, { "epoch": 1.531913192917949, "grad_norm": 0.1199939176440239, "learning_rate": 5.7557252301811284e-05, "loss": 0.0293, "step": 76660 }, { "epoch": 1.5321130250589505, "grad_norm": 0.11691582202911377, "learning_rate": 5.7547036677417407e-05, "loss": 0.0275, "step": 76670 }, { "epoch": 1.532312857199952, "grad_norm": 0.11020442098379135, "learning_rate": 5.753682073062798e-05, "loss": 0.0267, "step": 76680 }, { "epoch": 1.5325126893409537, "grad_norm": 0.14115846157073975, "learning_rate": 5.7526604461879406e-05, "loss": 0.0297, "step": 76690 }, { "epoch": 1.532712521481955, "grad_norm": 0.11950523406267166, "learning_rate": 5.751638787160813e-05, "loss": 0.0435, "step": 76700 }, { "epoch": 1.5329123536229567, "grad_norm": 0.14200378954410553, "learning_rate": 5.750617096025057e-05, "loss": 0.0245, "step": 76710 }, { "epoch": 1.5331121857639582, "grad_norm": 0.09600116312503815, "learning_rate": 5.749595372824318e-05, "loss": 0.0326, "step": 76720 }, { "epoch": 1.5333120179049597, "grad_norm": 0.1368386596441269, "learning_rate": 5.748573617602241e-05, "loss": 0.0267, "step": 76730 }, { "epoch": 1.5335118500459615, "grad_norm": 0.15751060843467712, "learning_rate": 5.7475518304024734e-05, "loss": 0.038, "step": 76740 }, { "epoch": 1.533711682186963, "grad_norm": 0.13204249739646912, "learning_rate": 5.746530011268666e-05, "loss": 0.0264, "step": 76750 }, { "epoch": 1.5339115143279645, "grad_norm": 0.11932744085788727, "learning_rate": 5.745508160244467e-05, "loss": 0.0356, "step": 76760 }, { "epoch": 1.5341113464689662, "grad_norm": 0.1333102583885193, "learning_rate": 5.7444862773735296e-05, "loss": 0.0383, "step": 76770 }, { "epoch": 1.5343111786099675, "grad_norm": 0.16211213171482086, "learning_rate": 5.743464362699506e-05, "loss": 0.0314, "step": 76780 }, { "epoch": 1.5345110107509692, "grad_norm": 0.16123852133750916, "learning_rate": 5.7424424162660505e-05, "loss": 0.0414, "step": 76790 }, { "epoch": 1.534710842891971, "grad_norm": 0.1320556104183197, "learning_rate": 5.74142043811682e-05, "loss": 0.0251, "step": 76800 }, { "epoch": 1.5349106750329722, "grad_norm": 0.14789707958698273, "learning_rate": 5.74039842829547e-05, "loss": 0.0252, "step": 76810 }, { "epoch": 1.535110507173974, "grad_norm": 0.13653266429901123, "learning_rate": 5.7393763868456584e-05, "loss": 0.0295, "step": 76820 }, { "epoch": 1.5353103393149754, "grad_norm": 0.20196260511875153, "learning_rate": 5.7383543138110465e-05, "loss": 0.0278, "step": 76830 }, { "epoch": 1.535510171455977, "grad_norm": 0.11615703254938126, "learning_rate": 5.7373322092352946e-05, "loss": 0.028, "step": 76840 }, { "epoch": 1.5357100035969786, "grad_norm": 0.11654571443796158, "learning_rate": 5.7363100731620656e-05, "loss": 0.0329, "step": 76850 }, { "epoch": 1.5359098357379801, "grad_norm": 0.12707683444023132, "learning_rate": 5.7352879056350216e-05, "loss": 0.0317, "step": 76860 }, { "epoch": 1.5361096678789816, "grad_norm": 0.13297462463378906, "learning_rate": 5.734265706697831e-05, "loss": 0.0285, "step": 76870 }, { "epoch": 1.5363095000199833, "grad_norm": 0.113243967294693, "learning_rate": 5.733243476394157e-05, "loss": 0.0374, "step": 76880 }, { "epoch": 1.5365093321609846, "grad_norm": 0.16657225787639618, "learning_rate": 5.732221214767668e-05, "loss": 0.0362, "step": 76890 }, { "epoch": 1.5367091643019863, "grad_norm": 0.12031858414411545, "learning_rate": 5.731198921862037e-05, "loss": 0.0316, "step": 76900 }, { "epoch": 1.5369089964429878, "grad_norm": 0.14982382953166962, "learning_rate": 5.7301765977209265e-05, "loss": 0.0282, "step": 76910 }, { "epoch": 1.5371088285839893, "grad_norm": 0.15503168106079102, "learning_rate": 5.7291542423880176e-05, "loss": 0.0312, "step": 76920 }, { "epoch": 1.537308660724991, "grad_norm": 0.1256619393825531, "learning_rate": 5.7281318559069764e-05, "loss": 0.0263, "step": 76930 }, { "epoch": 1.5375084928659926, "grad_norm": 0.10895868390798569, "learning_rate": 5.727109438321481e-05, "loss": 0.029, "step": 76940 }, { "epoch": 1.537708325006994, "grad_norm": 0.2078406661748886, "learning_rate": 5.726086989675205e-05, "loss": 0.0305, "step": 76950 }, { "epoch": 1.5379081571479958, "grad_norm": 0.11738287657499313, "learning_rate": 5.725064510011827e-05, "loss": 0.0368, "step": 76960 }, { "epoch": 1.5381079892889973, "grad_norm": 0.1277254819869995, "learning_rate": 5.724041999375025e-05, "loss": 0.0302, "step": 76970 }, { "epoch": 1.5383078214299988, "grad_norm": 0.11882554739713669, "learning_rate": 5.72301945780848e-05, "loss": 0.0313, "step": 76980 }, { "epoch": 1.5385076535710005, "grad_norm": 0.16176360845565796, "learning_rate": 5.721996885355872e-05, "loss": 0.0399, "step": 76990 }, { "epoch": 1.5387074857120018, "grad_norm": 0.12167531251907349, "learning_rate": 5.720974282060882e-05, "loss": 0.0393, "step": 77000 }, { "epoch": 1.5389073178530035, "grad_norm": 0.11744455993175507, "learning_rate": 5.7199516479671963e-05, "loss": 0.0332, "step": 77010 }, { "epoch": 1.539107149994005, "grad_norm": 0.20243823528289795, "learning_rate": 5.718928983118499e-05, "loss": 0.0273, "step": 77020 }, { "epoch": 1.5393069821350065, "grad_norm": 0.13099944591522217, "learning_rate": 5.717906287558475e-05, "loss": 0.0435, "step": 77030 }, { "epoch": 1.5395068142760082, "grad_norm": 0.11679882556200027, "learning_rate": 5.7168835613308134e-05, "loss": 0.0272, "step": 77040 }, { "epoch": 1.5397066464170097, "grad_norm": 0.12356823682785034, "learning_rate": 5.715860804479205e-05, "loss": 0.022, "step": 77050 }, { "epoch": 1.5399064785580112, "grad_norm": 0.11776940524578094, "learning_rate": 5.7148380170473345e-05, "loss": 0.0306, "step": 77060 }, { "epoch": 1.540106310699013, "grad_norm": 0.11924383789300919, "learning_rate": 5.7138151990789015e-05, "loss": 0.0219, "step": 77070 }, { "epoch": 1.5403061428400144, "grad_norm": 0.1054302230477333, "learning_rate": 5.712792350617592e-05, "loss": 0.0284, "step": 77080 }, { "epoch": 1.540505974981016, "grad_norm": 0.12988761067390442, "learning_rate": 5.711769471707102e-05, "loss": 0.037, "step": 77090 }, { "epoch": 1.5407058071220177, "grad_norm": 0.1407804936170578, "learning_rate": 5.71074656239113e-05, "loss": 0.0321, "step": 77100 }, { "epoch": 1.540905639263019, "grad_norm": 0.1307019591331482, "learning_rate": 5.709723622713368e-05, "loss": 0.0385, "step": 77110 }, { "epoch": 1.5411054714040207, "grad_norm": 0.1641581505537033, "learning_rate": 5.7087006527175194e-05, "loss": 0.0415, "step": 77120 }, { "epoch": 1.5413053035450222, "grad_norm": 0.1304692178964615, "learning_rate": 5.707677652447279e-05, "loss": 0.0337, "step": 77130 }, { "epoch": 1.5415051356860237, "grad_norm": 0.1527150720357895, "learning_rate": 5.706654621946349e-05, "loss": 0.0274, "step": 77140 }, { "epoch": 1.5417049678270254, "grad_norm": 0.1456267088651657, "learning_rate": 5.705631561258432e-05, "loss": 0.0264, "step": 77150 }, { "epoch": 1.5419047999680269, "grad_norm": 0.12715071439743042, "learning_rate": 5.704608470427232e-05, "loss": 0.0319, "step": 77160 }, { "epoch": 1.5421046321090284, "grad_norm": 0.11633782088756561, "learning_rate": 5.703585349496452e-05, "loss": 0.0277, "step": 77170 }, { "epoch": 1.54230446425003, "grad_norm": 0.13234256207942963, "learning_rate": 5.702562198509798e-05, "loss": 0.0372, "step": 77180 }, { "epoch": 1.5425042963910314, "grad_norm": 0.11279865354299545, "learning_rate": 5.7015390175109787e-05, "loss": 0.0338, "step": 77190 }, { "epoch": 1.542704128532033, "grad_norm": 0.14391769468784332, "learning_rate": 5.7005158065437e-05, "loss": 0.034, "step": 77200 }, { "epoch": 1.5429039606730348, "grad_norm": 0.1800553798675537, "learning_rate": 5.699492565651674e-05, "loss": 0.0343, "step": 77210 }, { "epoch": 1.543103792814036, "grad_norm": 0.12188226729631424, "learning_rate": 5.698469294878609e-05, "loss": 0.0342, "step": 77220 }, { "epoch": 1.5433036249550378, "grad_norm": 0.13325893878936768, "learning_rate": 5.69744599426822e-05, "loss": 0.0271, "step": 77230 }, { "epoch": 1.5435034570960393, "grad_norm": 0.12335590273141861, "learning_rate": 5.696422663864219e-05, "loss": 0.0266, "step": 77240 }, { "epoch": 1.5437032892370408, "grad_norm": 0.12140928953886032, "learning_rate": 5.6953993037103226e-05, "loss": 0.0401, "step": 77250 }, { "epoch": 1.5439031213780425, "grad_norm": 0.0949992835521698, "learning_rate": 5.6943759138502426e-05, "loss": 0.0283, "step": 77260 }, { "epoch": 1.544102953519044, "grad_norm": 0.13398513197898865, "learning_rate": 5.693352494327702e-05, "loss": 0.0359, "step": 77270 }, { "epoch": 1.5443027856600455, "grad_norm": 0.14179864525794983, "learning_rate": 5.6923290451864156e-05, "loss": 0.0305, "step": 77280 }, { "epoch": 1.5445026178010473, "grad_norm": 0.1166650652885437, "learning_rate": 5.6913055664701034e-05, "loss": 0.0294, "step": 77290 }, { "epoch": 1.5447024499420485, "grad_norm": 0.16102436184883118, "learning_rate": 5.690282058222489e-05, "loss": 0.0345, "step": 77300 }, { "epoch": 1.5449022820830502, "grad_norm": 0.12043177336454391, "learning_rate": 5.6892585204872906e-05, "loss": 0.024, "step": 77310 }, { "epoch": 1.5451021142240517, "grad_norm": 0.12195699661970139, "learning_rate": 5.688234953308238e-05, "loss": 0.0296, "step": 77320 }, { "epoch": 1.5453019463650532, "grad_norm": 0.10480588674545288, "learning_rate": 5.6872113567290494e-05, "loss": 0.0313, "step": 77330 }, { "epoch": 1.545501778506055, "grad_norm": 0.12594786286354065, "learning_rate": 5.6861877307934554e-05, "loss": 0.0284, "step": 77340 }, { "epoch": 1.5457016106470565, "grad_norm": 0.12734009325504303, "learning_rate": 5.685164075545182e-05, "loss": 0.0352, "step": 77350 }, { "epoch": 1.545901442788058, "grad_norm": 0.1561337560415268, "learning_rate": 5.684140391027958e-05, "loss": 0.0284, "step": 77360 }, { "epoch": 1.5461012749290597, "grad_norm": 0.14127059280872345, "learning_rate": 5.683116677285514e-05, "loss": 0.0357, "step": 77370 }, { "epoch": 1.5463011070700612, "grad_norm": 0.1326579749584198, "learning_rate": 5.6820929343615784e-05, "loss": 0.0258, "step": 77380 }, { "epoch": 1.5465009392110627, "grad_norm": 0.13076825439929962, "learning_rate": 5.681069162299888e-05, "loss": 0.0315, "step": 77390 }, { "epoch": 1.5467007713520644, "grad_norm": 0.12171930819749832, "learning_rate": 5.680045361144173e-05, "loss": 0.0253, "step": 77400 }, { "epoch": 1.5469006034930657, "grad_norm": 0.11013976484537125, "learning_rate": 5.679021530938169e-05, "loss": 0.0285, "step": 77410 }, { "epoch": 1.5471004356340674, "grad_norm": 0.12078480422496796, "learning_rate": 5.677997671725613e-05, "loss": 0.0328, "step": 77420 }, { "epoch": 1.547300267775069, "grad_norm": 0.11644898355007172, "learning_rate": 5.6769737835502424e-05, "loss": 0.0296, "step": 77430 }, { "epoch": 1.5475000999160704, "grad_norm": 0.1031704843044281, "learning_rate": 5.675949866455794e-05, "loss": 0.0239, "step": 77440 }, { "epoch": 1.5476999320570721, "grad_norm": 0.10461197793483734, "learning_rate": 5.674925920486012e-05, "loss": 0.0276, "step": 77450 }, { "epoch": 1.5478997641980736, "grad_norm": 0.10316494852304459, "learning_rate": 5.673901945684632e-05, "loss": 0.0313, "step": 77460 }, { "epoch": 1.5480995963390751, "grad_norm": 0.14324280619621277, "learning_rate": 5.6728779420954e-05, "loss": 0.0347, "step": 77470 }, { "epoch": 1.5482994284800768, "grad_norm": 0.12922784686088562, "learning_rate": 5.671853909762057e-05, "loss": 0.0411, "step": 77480 }, { "epoch": 1.5484992606210783, "grad_norm": 0.1534731388092041, "learning_rate": 5.670829848728351e-05, "loss": 0.023, "step": 77490 }, { "epoch": 1.5486990927620798, "grad_norm": 0.1283334493637085, "learning_rate": 5.669805759038026e-05, "loss": 0.0373, "step": 77500 }, { "epoch": 1.5488989249030816, "grad_norm": 0.11619316041469574, "learning_rate": 5.668781640734828e-05, "loss": 0.0314, "step": 77510 }, { "epoch": 1.5490987570440828, "grad_norm": 0.12114010751247406, "learning_rate": 5.667757493862509e-05, "loss": 0.0284, "step": 77520 }, { "epoch": 1.5492985891850846, "grad_norm": 0.14066246151924133, "learning_rate": 5.666733318464815e-05, "loss": 0.0344, "step": 77530 }, { "epoch": 1.549498421326086, "grad_norm": 0.11616940051317215, "learning_rate": 5.665709114585499e-05, "loss": 0.0344, "step": 77540 }, { "epoch": 1.5496982534670876, "grad_norm": 0.108817920088768, "learning_rate": 5.664684882268313e-05, "loss": 0.0238, "step": 77550 }, { "epoch": 1.5498980856080893, "grad_norm": 0.13097013533115387, "learning_rate": 5.6636606215570084e-05, "loss": 0.0358, "step": 77560 }, { "epoch": 1.5500979177490908, "grad_norm": 0.14126531779766083, "learning_rate": 5.662636332495343e-05, "loss": 0.0303, "step": 77570 }, { "epoch": 1.5502977498900923, "grad_norm": 0.13146857917308807, "learning_rate": 5.661612015127069e-05, "loss": 0.0405, "step": 77580 }, { "epoch": 1.550497582031094, "grad_norm": 0.1481773853302002, "learning_rate": 5.660587669495947e-05, "loss": 0.0285, "step": 77590 }, { "epoch": 1.5506974141720953, "grad_norm": 0.1192859411239624, "learning_rate": 5.6595632956457324e-05, "loss": 0.0274, "step": 77600 }, { "epoch": 1.550897246313097, "grad_norm": 0.11028771847486496, "learning_rate": 5.658538893620184e-05, "loss": 0.0241, "step": 77610 }, { "epoch": 1.5510970784540987, "grad_norm": 0.1171373724937439, "learning_rate": 5.657514463463066e-05, "loss": 0.0391, "step": 77620 }, { "epoch": 1.5512969105951, "grad_norm": 0.1531912088394165, "learning_rate": 5.656490005218137e-05, "loss": 0.0251, "step": 77630 }, { "epoch": 1.5514967427361017, "grad_norm": 0.11328206211328506, "learning_rate": 5.655465518929163e-05, "loss": 0.0379, "step": 77640 }, { "epoch": 1.5516965748771032, "grad_norm": 0.10758455842733383, "learning_rate": 5.654441004639903e-05, "loss": 0.0294, "step": 77650 }, { "epoch": 1.5518964070181047, "grad_norm": 0.11898475885391235, "learning_rate": 5.6534164623941274e-05, "loss": 0.0235, "step": 77660 }, { "epoch": 1.5520962391591064, "grad_norm": 0.13196389377117157, "learning_rate": 5.6523918922356e-05, "loss": 0.0338, "step": 77670 }, { "epoch": 1.552296071300108, "grad_norm": 0.21155239641666412, "learning_rate": 5.65136729420809e-05, "loss": 0.0324, "step": 77680 }, { "epoch": 1.5524959034411094, "grad_norm": 0.12354953587055206, "learning_rate": 5.650342668355366e-05, "loss": 0.023, "step": 77690 }, { "epoch": 1.5526957355821112, "grad_norm": 0.16402855515480042, "learning_rate": 5.649318014721199e-05, "loss": 0.0342, "step": 77700 }, { "epoch": 1.5528955677231124, "grad_norm": 0.14002852141857147, "learning_rate": 5.648293333349356e-05, "loss": 0.0275, "step": 77710 }, { "epoch": 1.5530953998641142, "grad_norm": 0.16980266571044922, "learning_rate": 5.647268624283616e-05, "loss": 0.0321, "step": 77720 }, { "epoch": 1.5532952320051157, "grad_norm": 0.15876223146915436, "learning_rate": 5.646243887567747e-05, "loss": 0.0393, "step": 77730 }, { "epoch": 1.5534950641461172, "grad_norm": 0.11731000989675522, "learning_rate": 5.6452191232455273e-05, "loss": 0.0328, "step": 77740 }, { "epoch": 1.5536948962871189, "grad_norm": 0.13180620968341827, "learning_rate": 5.644194331360734e-05, "loss": 0.0323, "step": 77750 }, { "epoch": 1.5538947284281204, "grad_norm": 0.2333376705646515, "learning_rate": 5.643169511957138e-05, "loss": 0.0347, "step": 77760 }, { "epoch": 1.5540945605691219, "grad_norm": 0.13961265981197357, "learning_rate": 5.642144665078526e-05, "loss": 0.0317, "step": 77770 }, { "epoch": 1.5542943927101236, "grad_norm": 0.1403249055147171, "learning_rate": 5.641119790768671e-05, "loss": 0.0233, "step": 77780 }, { "epoch": 1.554494224851125, "grad_norm": 0.10291747003793716, "learning_rate": 5.640094889071358e-05, "loss": 0.0245, "step": 77790 }, { "epoch": 1.5546940569921266, "grad_norm": 0.10529334843158722, "learning_rate": 5.6390699600303653e-05, "loss": 0.0233, "step": 77800 }, { "epoch": 1.5548938891331283, "grad_norm": 0.1552380621433258, "learning_rate": 5.638045003689479e-05, "loss": 0.0263, "step": 77810 }, { "epoch": 1.5550937212741296, "grad_norm": 0.11762506514787674, "learning_rate": 5.6370200200924825e-05, "loss": 0.0323, "step": 77820 }, { "epoch": 1.5552935534151313, "grad_norm": 0.13248656690120697, "learning_rate": 5.6359950092831605e-05, "loss": 0.0315, "step": 77830 }, { "epoch": 1.5554933855561328, "grad_norm": 0.13319659233093262, "learning_rate": 5.6349699713053014e-05, "loss": 0.0302, "step": 77840 }, { "epoch": 1.5556932176971343, "grad_norm": 0.14927710592746735, "learning_rate": 5.633944906202691e-05, "loss": 0.0337, "step": 77850 }, { "epoch": 1.555893049838136, "grad_norm": 0.10595019161701202, "learning_rate": 5.6329198140191173e-05, "loss": 0.024, "step": 77860 }, { "epoch": 1.5560928819791375, "grad_norm": 0.13352932035923004, "learning_rate": 5.631894694798373e-05, "loss": 0.0274, "step": 77870 }, { "epoch": 1.556292714120139, "grad_norm": 0.1422496885061264, "learning_rate": 5.630869548584248e-05, "loss": 0.0321, "step": 77880 }, { "epoch": 1.5564925462611408, "grad_norm": 0.15195332467556, "learning_rate": 5.6298443754205353e-05, "loss": 0.0281, "step": 77890 }, { "epoch": 1.5566923784021423, "grad_norm": 0.12678948044776917, "learning_rate": 5.628819175351028e-05, "loss": 0.0262, "step": 77900 }, { "epoch": 1.5568922105431438, "grad_norm": 0.11778657138347626, "learning_rate": 5.6277939484195195e-05, "loss": 0.0264, "step": 77910 }, { "epoch": 1.5570920426841455, "grad_norm": 0.12518025934696198, "learning_rate": 5.626768694669808e-05, "loss": 0.0383, "step": 77920 }, { "epoch": 1.5572918748251467, "grad_norm": 0.14906993508338928, "learning_rate": 5.625743414145689e-05, "loss": 0.0306, "step": 77930 }, { "epoch": 1.5574917069661485, "grad_norm": 0.12724778056144714, "learning_rate": 5.624718106890961e-05, "loss": 0.0244, "step": 77940 }, { "epoch": 1.55769153910715, "grad_norm": 0.1305728554725647, "learning_rate": 5.6236927729494235e-05, "loss": 0.0293, "step": 77950 }, { "epoch": 1.5578913712481515, "grad_norm": 0.1190590038895607, "learning_rate": 5.622667412364875e-05, "loss": 0.031, "step": 77960 }, { "epoch": 1.5580912033891532, "grad_norm": 0.13514283299446106, "learning_rate": 5.62164202518112e-05, "loss": 0.0265, "step": 77970 }, { "epoch": 1.5582910355301547, "grad_norm": 0.12028132379055023, "learning_rate": 5.620616611441958e-05, "loss": 0.0259, "step": 77980 }, { "epoch": 1.5584908676711562, "grad_norm": 0.11021532863378525, "learning_rate": 5.6195911711911966e-05, "loss": 0.034, "step": 77990 }, { "epoch": 1.558690699812158, "grad_norm": 0.13276050984859467, "learning_rate": 5.618565704472637e-05, "loss": 0.0246, "step": 78000 }, { "epoch": 1.5588905319531592, "grad_norm": 0.10001427680253983, "learning_rate": 5.6175402113300865e-05, "loss": 0.0314, "step": 78010 }, { "epoch": 1.559090364094161, "grad_norm": 0.12019342929124832, "learning_rate": 5.6165146918073554e-05, "loss": 0.0435, "step": 78020 }, { "epoch": 1.5592901962351626, "grad_norm": 0.137425035238266, "learning_rate": 5.6154891459482464e-05, "loss": 0.0397, "step": 78030 }, { "epoch": 1.559490028376164, "grad_norm": 0.11917918920516968, "learning_rate": 5.614463573796573e-05, "loss": 0.0303, "step": 78040 }, { "epoch": 1.5596898605171656, "grad_norm": 0.15028664469718933, "learning_rate": 5.613437975396144e-05, "loss": 0.0278, "step": 78050 }, { "epoch": 1.5598896926581671, "grad_norm": 0.11055401712656021, "learning_rate": 5.6124123507907725e-05, "loss": 0.028, "step": 78060 }, { "epoch": 1.5600895247991686, "grad_norm": 0.23993447422981262, "learning_rate": 5.611386700024269e-05, "loss": 0.0293, "step": 78070 }, { "epoch": 1.5602893569401703, "grad_norm": 0.12881913781166077, "learning_rate": 5.610361023140449e-05, "loss": 0.0356, "step": 78080 }, { "epoch": 1.5604891890811718, "grad_norm": 0.17175163328647614, "learning_rate": 5.609335320183128e-05, "loss": 0.042, "step": 78090 }, { "epoch": 1.5606890212221733, "grad_norm": 0.1546960026025772, "learning_rate": 5.6083095911961226e-05, "loss": 0.0307, "step": 78100 }, { "epoch": 1.560888853363175, "grad_norm": 0.15942883491516113, "learning_rate": 5.6072838362232475e-05, "loss": 0.0252, "step": 78110 }, { "epoch": 1.5610886855041763, "grad_norm": 0.1450299471616745, "learning_rate": 5.606258055308322e-05, "loss": 0.0328, "step": 78120 }, { "epoch": 1.561288517645178, "grad_norm": 0.1359458714723587, "learning_rate": 5.605232248495167e-05, "loss": 0.03, "step": 78130 }, { "epoch": 1.5614883497861796, "grad_norm": 0.13832609355449677, "learning_rate": 5.604206415827602e-05, "loss": 0.0319, "step": 78140 }, { "epoch": 1.561688181927181, "grad_norm": 0.13516300916671753, "learning_rate": 5.603180557349449e-05, "loss": 0.0252, "step": 78150 }, { "epoch": 1.5618880140681828, "grad_norm": 0.10746336728334427, "learning_rate": 5.60215467310453e-05, "loss": 0.0218, "step": 78160 }, { "epoch": 1.5620878462091843, "grad_norm": 0.16759322583675385, "learning_rate": 5.601128763136672e-05, "loss": 0.0255, "step": 78170 }, { "epoch": 1.5622876783501858, "grad_norm": 0.15009549260139465, "learning_rate": 5.600102827489695e-05, "loss": 0.0386, "step": 78180 }, { "epoch": 1.5624875104911875, "grad_norm": 0.12830503284931183, "learning_rate": 5.599076866207429e-05, "loss": 0.0282, "step": 78190 }, { "epoch": 1.562687342632189, "grad_norm": 0.14001356065273285, "learning_rate": 5.598050879333699e-05, "loss": 0.0324, "step": 78200 }, { "epoch": 1.5628871747731905, "grad_norm": 0.11644893139600754, "learning_rate": 5.5970248669123346e-05, "loss": 0.0371, "step": 78210 }, { "epoch": 1.5630870069141922, "grad_norm": 0.10985736548900604, "learning_rate": 5.5959988289871656e-05, "loss": 0.0291, "step": 78220 }, { "epoch": 1.5632868390551935, "grad_norm": 0.13211804628372192, "learning_rate": 5.594972765602019e-05, "loss": 0.0287, "step": 78230 }, { "epoch": 1.5634866711961952, "grad_norm": 0.11589514464139938, "learning_rate": 5.593946676800731e-05, "loss": 0.0255, "step": 78240 }, { "epoch": 1.5636865033371967, "grad_norm": 0.1677742451429367, "learning_rate": 5.5929205626271306e-05, "loss": 0.0328, "step": 78250 }, { "epoch": 1.5638863354781982, "grad_norm": 0.10580404847860336, "learning_rate": 5.5918944231250536e-05, "loss": 0.0262, "step": 78260 }, { "epoch": 1.5640861676192, "grad_norm": 0.16016823053359985, "learning_rate": 5.5908682583383344e-05, "loss": 0.0353, "step": 78270 }, { "epoch": 1.5642859997602014, "grad_norm": 0.11052737385034561, "learning_rate": 5.589842068310808e-05, "loss": 0.0369, "step": 78280 }, { "epoch": 1.564485831901203, "grad_norm": 0.12194187939167023, "learning_rate": 5.5888158530863124e-05, "loss": 0.0334, "step": 78290 }, { "epoch": 1.5646856640422047, "grad_norm": 0.13137952983379364, "learning_rate": 5.5877896127086856e-05, "loss": 0.0258, "step": 78300 }, { "epoch": 1.5648854961832062, "grad_norm": 0.11250977963209152, "learning_rate": 5.586763347221766e-05, "loss": 0.0292, "step": 78310 }, { "epoch": 1.5650853283242077, "grad_norm": 0.12471456825733185, "learning_rate": 5.5857370566693945e-05, "loss": 0.0317, "step": 78320 }, { "epoch": 1.5652851604652094, "grad_norm": 0.12854674458503723, "learning_rate": 5.5847107410954116e-05, "loss": 0.0286, "step": 78330 }, { "epoch": 1.5654849926062107, "grad_norm": 0.14634399116039276, "learning_rate": 5.5836844005436604e-05, "loss": 0.032, "step": 78340 }, { "epoch": 1.5656848247472124, "grad_norm": 0.10104206204414368, "learning_rate": 5.5826580350579826e-05, "loss": 0.0232, "step": 78350 }, { "epoch": 1.5658846568882139, "grad_norm": 0.12577742338180542, "learning_rate": 5.581631644682226e-05, "loss": 0.0292, "step": 78360 }, { "epoch": 1.5660844890292154, "grad_norm": 0.10759139060974121, "learning_rate": 5.580605229460233e-05, "loss": 0.0298, "step": 78370 }, { "epoch": 1.566284321170217, "grad_norm": 0.12111490964889526, "learning_rate": 5.57957878943585e-05, "loss": 0.0307, "step": 78380 }, { "epoch": 1.5664841533112186, "grad_norm": 0.14916111528873444, "learning_rate": 5.5785523246529284e-05, "loss": 0.0282, "step": 78390 }, { "epoch": 1.56668398545222, "grad_norm": 0.12370244413614273, "learning_rate": 5.577525835155313e-05, "loss": 0.0339, "step": 78400 }, { "epoch": 1.5668838175932218, "grad_norm": 0.14537440240383148, "learning_rate": 5.576499320986855e-05, "loss": 0.0297, "step": 78410 }, { "epoch": 1.567083649734223, "grad_norm": 0.18483169376850128, "learning_rate": 5.575472782191404e-05, "loss": 0.0282, "step": 78420 }, { "epoch": 1.5672834818752248, "grad_norm": 0.12976275384426117, "learning_rate": 5.5744462188128136e-05, "loss": 0.0322, "step": 78430 }, { "epoch": 1.5674833140162263, "grad_norm": 0.1427825540304184, "learning_rate": 5.5734196308949374e-05, "loss": 0.0339, "step": 78440 }, { "epoch": 1.5676831461572278, "grad_norm": 0.15071254968643188, "learning_rate": 5.572393018481625e-05, "loss": 0.0302, "step": 78450 }, { "epoch": 1.5678829782982295, "grad_norm": 0.11212427169084549, "learning_rate": 5.571366381616737e-05, "loss": 0.0241, "step": 78460 }, { "epoch": 1.568082810439231, "grad_norm": 0.11618469655513763, "learning_rate": 5.570339720344125e-05, "loss": 0.0337, "step": 78470 }, { "epoch": 1.5682826425802325, "grad_norm": 0.12189733237028122, "learning_rate": 5.569313034707649e-05, "loss": 0.0265, "step": 78480 }, { "epoch": 1.5684824747212343, "grad_norm": 0.13597089052200317, "learning_rate": 5.5682863247511674e-05, "loss": 0.0374, "step": 78490 }, { "epoch": 1.5686823068622358, "grad_norm": 0.29814353585243225, "learning_rate": 5.5672595905185356e-05, "loss": 0.0291, "step": 78500 }, { "epoch": 1.5688821390032373, "grad_norm": 0.16031989455223083, "learning_rate": 5.5662328320536174e-05, "loss": 0.0333, "step": 78510 }, { "epoch": 1.569081971144239, "grad_norm": 0.12866553664207458, "learning_rate": 5.565206049400272e-05, "loss": 0.0255, "step": 78520 }, { "epoch": 1.5692818032852403, "grad_norm": 0.14655806124210358, "learning_rate": 5.564179242602363e-05, "loss": 0.0273, "step": 78530 }, { "epoch": 1.569481635426242, "grad_norm": 0.1123352125287056, "learning_rate": 5.563152411703753e-05, "loss": 0.0262, "step": 78540 }, { "epoch": 1.5696814675672435, "grad_norm": 0.13943517208099365, "learning_rate": 5.562125556748306e-05, "loss": 0.0323, "step": 78550 }, { "epoch": 1.569881299708245, "grad_norm": 0.1062173992395401, "learning_rate": 5.561098677779888e-05, "loss": 0.0245, "step": 78560 }, { "epoch": 1.5700811318492467, "grad_norm": 0.1293983906507492, "learning_rate": 5.5600717748423666e-05, "loss": 0.0293, "step": 78570 }, { "epoch": 1.5702809639902482, "grad_norm": 0.19245636463165283, "learning_rate": 5.5590448479796074e-05, "loss": 0.0364, "step": 78580 }, { "epoch": 1.5704807961312497, "grad_norm": 0.14998477697372437, "learning_rate": 5.558017897235479e-05, "loss": 0.0294, "step": 78590 }, { "epoch": 1.5706806282722514, "grad_norm": 0.19098906219005585, "learning_rate": 5.556990922653851e-05, "loss": 0.0269, "step": 78600 }, { "epoch": 1.570880460413253, "grad_norm": 0.13752378523349762, "learning_rate": 5.555963924278593e-05, "loss": 0.0337, "step": 78610 }, { "epoch": 1.5710802925542544, "grad_norm": 0.15627743303775787, "learning_rate": 5.55493690215358e-05, "loss": 0.0352, "step": 78620 }, { "epoch": 1.5712801246952561, "grad_norm": 0.1640596091747284, "learning_rate": 5.5539098563226785e-05, "loss": 0.0311, "step": 78630 }, { "epoch": 1.5714799568362574, "grad_norm": 0.13040460646152496, "learning_rate": 5.552882786829769e-05, "loss": 0.0326, "step": 78640 }, { "epoch": 1.5716797889772591, "grad_norm": 0.146403968334198, "learning_rate": 5.551855693718722e-05, "loss": 0.0308, "step": 78650 }, { "epoch": 1.5718796211182606, "grad_norm": 0.12774768471717834, "learning_rate": 5.550828577033411e-05, "loss": 0.0275, "step": 78660 }, { "epoch": 1.5720794532592621, "grad_norm": 0.1608291119337082, "learning_rate": 5.549801436817719e-05, "loss": 0.0432, "step": 78670 }, { "epoch": 1.5722792854002638, "grad_norm": 0.1399621218442917, "learning_rate": 5.5487742731155156e-05, "loss": 0.0415, "step": 78680 }, { "epoch": 1.5724791175412653, "grad_norm": 0.15648360550403595, "learning_rate": 5.547747085970686e-05, "loss": 0.0265, "step": 78690 }, { "epoch": 1.5726789496822668, "grad_norm": 0.1419396996498108, "learning_rate": 5.546719875427105e-05, "loss": 0.0314, "step": 78700 }, { "epoch": 1.5728787818232686, "grad_norm": 0.10288970917463303, "learning_rate": 5.545692641528657e-05, "loss": 0.0248, "step": 78710 }, { "epoch": 1.57307861396427, "grad_norm": 0.1345260888338089, "learning_rate": 5.544665384319221e-05, "loss": 0.0332, "step": 78720 }, { "epoch": 1.5732784461052716, "grad_norm": 0.10374323278665543, "learning_rate": 5.543638103842681e-05, "loss": 0.0242, "step": 78730 }, { "epoch": 1.5734782782462733, "grad_norm": 0.12550212442874908, "learning_rate": 5.542610800142919e-05, "loss": 0.0245, "step": 78740 }, { "epoch": 1.5736781103872746, "grad_norm": 0.12709397077560425, "learning_rate": 5.541583473263822e-05, "loss": 0.0291, "step": 78750 }, { "epoch": 1.5738779425282763, "grad_norm": 0.13353660702705383, "learning_rate": 5.540556123249273e-05, "loss": 0.0306, "step": 78760 }, { "epoch": 1.5740777746692778, "grad_norm": 0.1351029872894287, "learning_rate": 5.539528750143159e-05, "loss": 0.0262, "step": 78770 }, { "epoch": 1.5742776068102793, "grad_norm": 0.11199343204498291, "learning_rate": 5.538501353989368e-05, "loss": 0.0293, "step": 78780 }, { "epoch": 1.574477438951281, "grad_norm": 0.1346714049577713, "learning_rate": 5.5374739348317895e-05, "loss": 0.021, "step": 78790 }, { "epoch": 1.5746772710922825, "grad_norm": 0.14666453003883362, "learning_rate": 5.536446492714311e-05, "loss": 0.0376, "step": 78800 }, { "epoch": 1.574877103233284, "grad_norm": 0.14421145617961884, "learning_rate": 5.535419027680824e-05, "loss": 0.0298, "step": 78810 }, { "epoch": 1.5750769353742857, "grad_norm": 0.1191955953836441, "learning_rate": 5.53439153977522e-05, "loss": 0.0264, "step": 78820 }, { "epoch": 1.575276767515287, "grad_norm": 0.1475783884525299, "learning_rate": 5.5333640290413904e-05, "loss": 0.0255, "step": 78830 }, { "epoch": 1.5754765996562887, "grad_norm": 0.16895101964473724, "learning_rate": 5.532336495523231e-05, "loss": 0.023, "step": 78840 }, { "epoch": 1.5756764317972902, "grad_norm": 0.1330907940864563, "learning_rate": 5.5313089392646325e-05, "loss": 0.0268, "step": 78850 }, { "epoch": 1.5758762639382917, "grad_norm": 0.13325375318527222, "learning_rate": 5.530281360309493e-05, "loss": 0.032, "step": 78860 }, { "epoch": 1.5760760960792934, "grad_norm": 0.18029233813285828, "learning_rate": 5.5292537587017076e-05, "loss": 0.0262, "step": 78870 }, { "epoch": 1.576275928220295, "grad_norm": 0.12846922874450684, "learning_rate": 5.5282261344851735e-05, "loss": 0.0368, "step": 78880 }, { "epoch": 1.5764757603612964, "grad_norm": 0.12327655404806137, "learning_rate": 5.527198487703791e-05, "loss": 0.0302, "step": 78890 }, { "epoch": 1.5766755925022982, "grad_norm": 0.1338949352502823, "learning_rate": 5.526170818401456e-05, "loss": 0.0299, "step": 78900 }, { "epoch": 1.5768754246432997, "grad_norm": 0.15040744841098785, "learning_rate": 5.525143126622071e-05, "loss": 0.0312, "step": 78910 }, { "epoch": 1.5770752567843012, "grad_norm": 0.12496823817491531, "learning_rate": 5.5241154124095365e-05, "loss": 0.0269, "step": 78920 }, { "epoch": 1.5772750889253029, "grad_norm": 0.10910478234291077, "learning_rate": 5.5230876758077535e-05, "loss": 0.0267, "step": 78930 }, { "epoch": 1.5774749210663042, "grad_norm": 0.1159321516752243, "learning_rate": 5.522059916860626e-05, "loss": 0.0318, "step": 78940 }, { "epoch": 1.5776747532073059, "grad_norm": 0.14685046672821045, "learning_rate": 5.5210321356120576e-05, "loss": 0.034, "step": 78950 }, { "epoch": 1.5778745853483074, "grad_norm": 0.09749916940927505, "learning_rate": 5.5200043321059556e-05, "loss": 0.0289, "step": 78960 }, { "epoch": 1.5780744174893089, "grad_norm": 0.159377783536911, "learning_rate": 5.518976506386222e-05, "loss": 0.0298, "step": 78970 }, { "epoch": 1.5782742496303106, "grad_norm": 0.12596087157726288, "learning_rate": 5.5179486584967666e-05, "loss": 0.0335, "step": 78980 }, { "epoch": 1.578474081771312, "grad_norm": 0.12129584699869156, "learning_rate": 5.5169207884814946e-05, "loss": 0.0345, "step": 78990 }, { "epoch": 1.5786739139123136, "grad_norm": 0.08469124138355255, "learning_rate": 5.515892896384317e-05, "loss": 0.0245, "step": 79000 }, { "epoch": 1.5788737460533153, "grad_norm": 0.12201303988695145, "learning_rate": 5.5148649822491424e-05, "loss": 0.0381, "step": 79010 }, { "epoch": 1.5790735781943168, "grad_norm": 0.14385153353214264, "learning_rate": 5.5138370461198827e-05, "loss": 0.0263, "step": 79020 }, { "epoch": 1.5792734103353183, "grad_norm": 0.1832522451877594, "learning_rate": 5.512809088040446e-05, "loss": 0.0314, "step": 79030 }, { "epoch": 1.57947324247632, "grad_norm": 0.12908217310905457, "learning_rate": 5.5117811080547486e-05, "loss": 0.0298, "step": 79040 }, { "epoch": 1.5796730746173213, "grad_norm": 0.1101442277431488, "learning_rate": 5.510753106206703e-05, "loss": 0.0327, "step": 79050 }, { "epoch": 1.579872906758323, "grad_norm": 0.12340071052312851, "learning_rate": 5.5097250825402214e-05, "loss": 0.0249, "step": 79060 }, { "epoch": 1.5800727388993245, "grad_norm": 0.11215948313474655, "learning_rate": 5.508697037099223e-05, "loss": 0.0293, "step": 79070 }, { "epoch": 1.580272571040326, "grad_norm": 0.13456325232982635, "learning_rate": 5.5076689699276185e-05, "loss": 0.0379, "step": 79080 }, { "epoch": 1.5804724031813278, "grad_norm": 0.11493184417486191, "learning_rate": 5.5066408810693317e-05, "loss": 0.029, "step": 79090 }, { "epoch": 1.5806722353223293, "grad_norm": 0.17176859080791473, "learning_rate": 5.505612770568275e-05, "loss": 0.0302, "step": 79100 }, { "epoch": 1.5808720674633308, "grad_norm": 0.12031958997249603, "learning_rate": 5.50458463846837e-05, "loss": 0.0278, "step": 79110 }, { "epoch": 1.5810718996043325, "grad_norm": 0.11619962006807327, "learning_rate": 5.503556484813536e-05, "loss": 0.0266, "step": 79120 }, { "epoch": 1.581271731745334, "grad_norm": 0.1152353510260582, "learning_rate": 5.5025283096476934e-05, "loss": 0.0371, "step": 79130 }, { "epoch": 1.5814715638863355, "grad_norm": 0.1403193473815918, "learning_rate": 5.5015001130147666e-05, "loss": 0.0385, "step": 79140 }, { "epoch": 1.5816713960273372, "grad_norm": 0.10332182794809341, "learning_rate": 5.5004718949586744e-05, "loss": 0.032, "step": 79150 }, { "epoch": 1.5818712281683385, "grad_norm": 0.15021535754203796, "learning_rate": 5.499443655523343e-05, "loss": 0.0274, "step": 79160 }, { "epoch": 1.5820710603093402, "grad_norm": 0.11523677408695221, "learning_rate": 5.498415394752696e-05, "loss": 0.0383, "step": 79170 }, { "epoch": 1.5822708924503417, "grad_norm": 0.09744907170534134, "learning_rate": 5.4973871126906596e-05, "loss": 0.0288, "step": 79180 }, { "epoch": 1.5824707245913432, "grad_norm": 0.14920607209205627, "learning_rate": 5.496358809381158e-05, "loss": 0.0343, "step": 79190 }, { "epoch": 1.582670556732345, "grad_norm": 0.1230136826634407, "learning_rate": 5.495330484868121e-05, "loss": 0.0249, "step": 79200 }, { "epoch": 1.5828703888733464, "grad_norm": 0.1394977569580078, "learning_rate": 5.4943021391954755e-05, "loss": 0.0348, "step": 79210 }, { "epoch": 1.583070221014348, "grad_norm": 0.15424229204654694, "learning_rate": 5.493273772407152e-05, "loss": 0.0336, "step": 79220 }, { "epoch": 1.5832700531553496, "grad_norm": 0.12340058386325836, "learning_rate": 5.492245384547077e-05, "loss": 0.0257, "step": 79230 }, { "epoch": 1.583469885296351, "grad_norm": 0.16137604415416718, "learning_rate": 5.491216975659184e-05, "loss": 0.0293, "step": 79240 }, { "epoch": 1.5836697174373526, "grad_norm": 0.12504325807094574, "learning_rate": 5.490188545787404e-05, "loss": 0.0288, "step": 79250 }, { "epoch": 1.5838695495783541, "grad_norm": 0.11719997972249985, "learning_rate": 5.489160094975669e-05, "loss": 0.0294, "step": 79260 }, { "epoch": 1.5840693817193556, "grad_norm": 0.1263451874256134, "learning_rate": 5.488131623267916e-05, "loss": 0.0243, "step": 79270 }, { "epoch": 1.5842692138603574, "grad_norm": 0.13134585320949554, "learning_rate": 5.4871031307080745e-05, "loss": 0.0318, "step": 79280 }, { "epoch": 1.5844690460013588, "grad_norm": 0.13674740493297577, "learning_rate": 5.486074617340083e-05, "loss": 0.0281, "step": 79290 }, { "epoch": 1.5846688781423603, "grad_norm": 0.11859377473592758, "learning_rate": 5.4850460832078774e-05, "loss": 0.031, "step": 79300 }, { "epoch": 1.584868710283362, "grad_norm": 0.10871140658855438, "learning_rate": 5.484017528355393e-05, "loss": 0.0408, "step": 79310 }, { "epoch": 1.5850685424243636, "grad_norm": 0.14219066500663757, "learning_rate": 5.482988952826571e-05, "loss": 0.0265, "step": 79320 }, { "epoch": 1.585268374565365, "grad_norm": 0.16189391911029816, "learning_rate": 5.4819603566653455e-05, "loss": 0.0278, "step": 79330 }, { "epoch": 1.5854682067063668, "grad_norm": 0.14886991679668427, "learning_rate": 5.48093173991566e-05, "loss": 0.0305, "step": 79340 }, { "epoch": 1.585668038847368, "grad_norm": 0.14972294867038727, "learning_rate": 5.479903102621453e-05, "loss": 0.0286, "step": 79350 }, { "epoch": 1.5858678709883698, "grad_norm": 0.11491622030735016, "learning_rate": 5.4788744448266695e-05, "loss": 0.0315, "step": 79360 }, { "epoch": 1.5860677031293713, "grad_norm": 0.15024305880069733, "learning_rate": 5.477845766575248e-05, "loss": 0.03, "step": 79370 }, { "epoch": 1.5862675352703728, "grad_norm": 0.1261330395936966, "learning_rate": 5.476817067911133e-05, "loss": 0.0303, "step": 79380 }, { "epoch": 1.5864673674113745, "grad_norm": 0.13638342916965485, "learning_rate": 5.47578834887827e-05, "loss": 0.0326, "step": 79390 }, { "epoch": 1.586667199552376, "grad_norm": 0.12246726453304291, "learning_rate": 5.474759609520601e-05, "loss": 0.0268, "step": 79400 }, { "epoch": 1.5868670316933775, "grad_norm": 0.1371426284313202, "learning_rate": 5.473730849882076e-05, "loss": 0.0343, "step": 79410 }, { "epoch": 1.5870668638343792, "grad_norm": 0.1211409643292427, "learning_rate": 5.4727020700066376e-05, "loss": 0.0346, "step": 79420 }, { "epoch": 1.5872666959753807, "grad_norm": 0.12152861803770065, "learning_rate": 5.471673269938236e-05, "loss": 0.0251, "step": 79430 }, { "epoch": 1.5874665281163822, "grad_norm": 0.1089133769273758, "learning_rate": 5.470644449720817e-05, "loss": 0.0342, "step": 79440 }, { "epoch": 1.587666360257384, "grad_norm": 0.18631723523139954, "learning_rate": 5.469615609398333e-05, "loss": 0.031, "step": 79450 }, { "epoch": 1.5878661923983852, "grad_norm": 0.15486928820610046, "learning_rate": 5.468586749014734e-05, "loss": 0.0334, "step": 79460 }, { "epoch": 1.588066024539387, "grad_norm": 0.12935537099838257, "learning_rate": 5.467557868613969e-05, "loss": 0.0335, "step": 79470 }, { "epoch": 1.5882658566803884, "grad_norm": 0.13858939707279205, "learning_rate": 5.46652896823999e-05, "loss": 0.0252, "step": 79480 }, { "epoch": 1.58846568882139, "grad_norm": 0.1367574781179428, "learning_rate": 5.465500047936751e-05, "loss": 0.0338, "step": 79490 }, { "epoch": 1.5886655209623917, "grad_norm": 0.11418640613555908, "learning_rate": 5.464471107748206e-05, "loss": 0.0283, "step": 79500 }, { "epoch": 1.5888653531033932, "grad_norm": 0.08968667685985565, "learning_rate": 5.4634421477183076e-05, "loss": 0.0266, "step": 79510 }, { "epoch": 1.5890651852443947, "grad_norm": 0.15444837510585785, "learning_rate": 5.462413167891013e-05, "loss": 0.0354, "step": 79520 }, { "epoch": 1.5892650173853964, "grad_norm": 0.1615547239780426, "learning_rate": 5.461384168310275e-05, "loss": 0.0295, "step": 79530 }, { "epoch": 1.5894648495263977, "grad_norm": 0.1539057344198227, "learning_rate": 5.460355149020056e-05, "loss": 0.0297, "step": 79540 }, { "epoch": 1.5896646816673994, "grad_norm": 0.08688493818044662, "learning_rate": 5.459326110064308e-05, "loss": 0.0265, "step": 79550 }, { "epoch": 1.589864513808401, "grad_norm": 0.13704396784305573, "learning_rate": 5.458297051486993e-05, "loss": 0.0378, "step": 79560 }, { "epoch": 1.5900643459494024, "grad_norm": 0.12905125319957733, "learning_rate": 5.457267973332071e-05, "loss": 0.0334, "step": 79570 }, { "epoch": 1.590264178090404, "grad_norm": 0.1150646060705185, "learning_rate": 5.4562388756435e-05, "loss": 0.0317, "step": 79580 }, { "epoch": 1.5904640102314056, "grad_norm": 0.1197797879576683, "learning_rate": 5.455209758465244e-05, "loss": 0.0255, "step": 79590 }, { "epoch": 1.590663842372407, "grad_norm": 0.12679779529571533, "learning_rate": 5.454180621841262e-05, "loss": 0.0267, "step": 79600 }, { "epoch": 1.5908636745134088, "grad_norm": 0.12140250205993652, "learning_rate": 5.4531514658155194e-05, "loss": 0.0374, "step": 79610 }, { "epoch": 1.5910635066544103, "grad_norm": 0.12892507016658783, "learning_rate": 5.452122290431978e-05, "loss": 0.0416, "step": 79620 }, { "epoch": 1.5912633387954118, "grad_norm": 0.1175786554813385, "learning_rate": 5.451093095734602e-05, "loss": 0.0262, "step": 79630 }, { "epoch": 1.5914631709364135, "grad_norm": 0.12350155413150787, "learning_rate": 5.450063881767359e-05, "loss": 0.0298, "step": 79640 }, { "epoch": 1.5916630030774148, "grad_norm": 0.1308048814535141, "learning_rate": 5.4490346485742124e-05, "loss": 0.0289, "step": 79650 }, { "epoch": 1.5918628352184165, "grad_norm": 0.11987771838903427, "learning_rate": 5.448005396199131e-05, "loss": 0.027, "step": 79660 }, { "epoch": 1.592062667359418, "grad_norm": 0.10648126900196075, "learning_rate": 5.446976124686083e-05, "loss": 0.0303, "step": 79670 }, { "epoch": 1.5922624995004195, "grad_norm": 0.15198083221912384, "learning_rate": 5.4459468340790345e-05, "loss": 0.0318, "step": 79680 }, { "epoch": 1.5924623316414213, "grad_norm": 0.13213925063610077, "learning_rate": 5.444917524421959e-05, "loss": 0.0328, "step": 79690 }, { "epoch": 1.5926621637824228, "grad_norm": 0.16157487034797668, "learning_rate": 5.4438881957588216e-05, "loss": 0.0389, "step": 79700 }, { "epoch": 1.5928619959234243, "grad_norm": 0.14763005077838898, "learning_rate": 5.4428588481335976e-05, "loss": 0.0336, "step": 79710 }, { "epoch": 1.593061828064426, "grad_norm": 0.1469816118478775, "learning_rate": 5.441829481590257e-05, "loss": 0.0374, "step": 79720 }, { "epoch": 1.5932616602054275, "grad_norm": 0.14664794504642487, "learning_rate": 5.440800096172771e-05, "loss": 0.0306, "step": 79730 }, { "epoch": 1.593461492346429, "grad_norm": 0.2017296701669693, "learning_rate": 5.4397706919251165e-05, "loss": 0.0311, "step": 79740 }, { "epoch": 1.5936613244874307, "grad_norm": 0.13796444237232208, "learning_rate": 5.4387412688912645e-05, "loss": 0.0282, "step": 79750 }, { "epoch": 1.593861156628432, "grad_norm": 0.11939218640327454, "learning_rate": 5.437711827115193e-05, "loss": 0.024, "step": 79760 }, { "epoch": 1.5940609887694337, "grad_norm": 0.11013053357601166, "learning_rate": 5.4366823666408754e-05, "loss": 0.033, "step": 79770 }, { "epoch": 1.5942608209104352, "grad_norm": 0.13857144117355347, "learning_rate": 5.4356528875122895e-05, "loss": 0.0251, "step": 79780 }, { "epoch": 1.5944606530514367, "grad_norm": 0.12904687225818634, "learning_rate": 5.4346233897734125e-05, "loss": 0.0282, "step": 79790 }, { "epoch": 1.5946604851924384, "grad_norm": 0.11709519475698471, "learning_rate": 5.433593873468221e-05, "loss": 0.0355, "step": 79800 }, { "epoch": 1.59486031733344, "grad_norm": 0.1288810670375824, "learning_rate": 5.432564338640699e-05, "loss": 0.0387, "step": 79810 }, { "epoch": 1.5950601494744414, "grad_norm": 0.0929679274559021, "learning_rate": 5.431534785334822e-05, "loss": 0.0272, "step": 79820 }, { "epoch": 1.5952599816154431, "grad_norm": 0.1154838427901268, "learning_rate": 5.430505213594571e-05, "loss": 0.0313, "step": 79830 }, { "epoch": 1.5954598137564446, "grad_norm": 0.14051155745983124, "learning_rate": 5.4294756234639274e-05, "loss": 0.0434, "step": 79840 }, { "epoch": 1.5956596458974461, "grad_norm": 0.1171039491891861, "learning_rate": 5.4284460149868754e-05, "loss": 0.0287, "step": 79850 }, { "epoch": 1.5958594780384479, "grad_norm": 0.1100655123591423, "learning_rate": 5.427416388207397e-05, "loss": 0.022, "step": 79860 }, { "epoch": 1.5960593101794491, "grad_norm": 0.15641826391220093, "learning_rate": 5.426386743169476e-05, "loss": 0.0274, "step": 79870 }, { "epoch": 1.5962591423204509, "grad_norm": 0.1362607628107071, "learning_rate": 5.425357079917095e-05, "loss": 0.0312, "step": 79880 }, { "epoch": 1.5964589744614524, "grad_norm": 0.10750263184309006, "learning_rate": 5.424327398494242e-05, "loss": 0.0367, "step": 79890 }, { "epoch": 1.5966588066024539, "grad_norm": 0.09816686064004898, "learning_rate": 5.4232976989449016e-05, "loss": 0.0241, "step": 79900 }, { "epoch": 1.5968586387434556, "grad_norm": 0.13136935234069824, "learning_rate": 5.422267981313062e-05, "loss": 0.0294, "step": 79910 }, { "epoch": 1.597058470884457, "grad_norm": 0.14592371881008148, "learning_rate": 5.42123824564271e-05, "loss": 0.0296, "step": 79920 }, { "epoch": 1.5972583030254586, "grad_norm": 0.12950028479099274, "learning_rate": 5.4202084919778316e-05, "loss": 0.0212, "step": 79930 }, { "epoch": 1.5974581351664603, "grad_norm": 0.15865658223628998, "learning_rate": 5.419178720362421e-05, "loss": 0.0257, "step": 79940 }, { "epoch": 1.5976579673074616, "grad_norm": 0.10763434320688248, "learning_rate": 5.4181489308404635e-05, "loss": 0.0318, "step": 79950 }, { "epoch": 1.5978577994484633, "grad_norm": 0.15585385262966156, "learning_rate": 5.417119123455954e-05, "loss": 0.0274, "step": 79960 }, { "epoch": 1.598057631589465, "grad_norm": 0.1394006460905075, "learning_rate": 5.41608929825288e-05, "loss": 0.0351, "step": 79970 }, { "epoch": 1.5982574637304663, "grad_norm": 0.1322292536497116, "learning_rate": 5.415059455275236e-05, "loss": 0.0285, "step": 79980 }, { "epoch": 1.598457295871468, "grad_norm": 0.14005373418331146, "learning_rate": 5.414029594567016e-05, "loss": 0.0275, "step": 79990 }, { "epoch": 1.5986571280124695, "grad_norm": 0.14657095074653625, "learning_rate": 5.41299971617221e-05, "loss": 0.0349, "step": 80000 }, { "epoch": 1.598856960153471, "grad_norm": 0.16692763566970825, "learning_rate": 5.4119698201348165e-05, "loss": 0.0275, "step": 80010 }, { "epoch": 1.5990567922944727, "grad_norm": 0.1303129941225052, "learning_rate": 5.4109399064988296e-05, "loss": 0.0252, "step": 80020 }, { "epoch": 1.5992566244354742, "grad_norm": 0.38707995414733887, "learning_rate": 5.409909975308244e-05, "loss": 0.0351, "step": 80030 }, { "epoch": 1.5994564565764757, "grad_norm": 0.10996674001216888, "learning_rate": 5.408880026607056e-05, "loss": 0.0201, "step": 80040 }, { "epoch": 1.5996562887174774, "grad_norm": 0.1448931246995926, "learning_rate": 5.4078500604392656e-05, "loss": 0.0365, "step": 80050 }, { "epoch": 1.5998561208584787, "grad_norm": 0.11330819875001907, "learning_rate": 5.40682007684887e-05, "loss": 0.0258, "step": 80060 }, { "epoch": 1.6000559529994804, "grad_norm": 0.14157965779304504, "learning_rate": 5.405790075879869e-05, "loss": 0.0298, "step": 80070 }, { "epoch": 1.600255785140482, "grad_norm": 0.2169024646282196, "learning_rate": 5.404760057576259e-05, "loss": 0.0262, "step": 80080 }, { "epoch": 1.6004556172814834, "grad_norm": 0.13239307701587677, "learning_rate": 5.403730021982045e-05, "loss": 0.0301, "step": 80090 }, { "epoch": 1.6006554494224852, "grad_norm": 0.11078637838363647, "learning_rate": 5.402699969141225e-05, "loss": 0.0328, "step": 80100 }, { "epoch": 1.6008552815634867, "grad_norm": 0.12557397782802582, "learning_rate": 5.401669899097802e-05, "loss": 0.0343, "step": 80110 }, { "epoch": 1.6010551137044882, "grad_norm": 0.12327650934457779, "learning_rate": 5.40063981189578e-05, "loss": 0.0285, "step": 80120 }, { "epoch": 1.6012549458454899, "grad_norm": 0.12196896225214005, "learning_rate": 5.3996097075791595e-05, "loss": 0.0346, "step": 80130 }, { "epoch": 1.6014547779864914, "grad_norm": 0.1261974722146988, "learning_rate": 5.3985795861919484e-05, "loss": 0.0264, "step": 80140 }, { "epoch": 1.6016546101274929, "grad_norm": 0.11502747237682343, "learning_rate": 5.397549447778147e-05, "loss": 0.0305, "step": 80150 }, { "epoch": 1.6018544422684946, "grad_norm": 0.117317333817482, "learning_rate": 5.3965192923817664e-05, "loss": 0.028, "step": 80160 }, { "epoch": 1.6020542744094959, "grad_norm": 0.13175274431705475, "learning_rate": 5.395489120046808e-05, "loss": 0.0311, "step": 80170 }, { "epoch": 1.6022541065504976, "grad_norm": 0.13818775117397308, "learning_rate": 5.394458930817282e-05, "loss": 0.0254, "step": 80180 }, { "epoch": 1.602453938691499, "grad_norm": 0.09794396162033081, "learning_rate": 5.393428724737195e-05, "loss": 0.0264, "step": 80190 }, { "epoch": 1.6026537708325006, "grad_norm": 0.14125171303749084, "learning_rate": 5.392398501850555e-05, "loss": 0.0277, "step": 80200 }, { "epoch": 1.6028536029735023, "grad_norm": 0.12081295251846313, "learning_rate": 5.391368262201373e-05, "loss": 0.0226, "step": 80210 }, { "epoch": 1.6030534351145038, "grad_norm": 0.10092002898454666, "learning_rate": 5.390338005833657e-05, "loss": 0.026, "step": 80220 }, { "epoch": 1.6032532672555053, "grad_norm": 0.1255664825439453, "learning_rate": 5.389307732791417e-05, "loss": 0.0338, "step": 80230 }, { "epoch": 1.603453099396507, "grad_norm": 0.18008224666118622, "learning_rate": 5.3882774431186676e-05, "loss": 0.0275, "step": 80240 }, { "epoch": 1.6036529315375085, "grad_norm": 0.09675724804401398, "learning_rate": 5.387247136859419e-05, "loss": 0.0287, "step": 80250 }, { "epoch": 1.60385276367851, "grad_norm": 0.08158716559410095, "learning_rate": 5.3862168140576854e-05, "loss": 0.0259, "step": 80260 }, { "epoch": 1.6040525958195118, "grad_norm": 0.13209021091461182, "learning_rate": 5.385186474757478e-05, "loss": 0.0354, "step": 80270 }, { "epoch": 1.604252427960513, "grad_norm": 0.09791669994592667, "learning_rate": 5.384156119002811e-05, "loss": 0.0353, "step": 80280 }, { "epoch": 1.6044522601015148, "grad_norm": 0.09533391147851944, "learning_rate": 5.383125746837701e-05, "loss": 0.0281, "step": 80290 }, { "epoch": 1.6046520922425163, "grad_norm": 0.11639029532670975, "learning_rate": 5.382095358306163e-05, "loss": 0.0307, "step": 80300 }, { "epoch": 1.6048519243835178, "grad_norm": 0.12373168766498566, "learning_rate": 5.381064953452212e-05, "loss": 0.0284, "step": 80310 }, { "epoch": 1.6050517565245195, "grad_norm": 0.14053913950920105, "learning_rate": 5.3800345323198686e-05, "loss": 0.0342, "step": 80320 }, { "epoch": 1.605251588665521, "grad_norm": 0.13035155832767487, "learning_rate": 5.379004094953145e-05, "loss": 0.0386, "step": 80330 }, { "epoch": 1.6054514208065225, "grad_norm": 0.09680639207363129, "learning_rate": 5.3779736413960656e-05, "loss": 0.0272, "step": 80340 }, { "epoch": 1.6056512529475242, "grad_norm": 0.11390335112810135, "learning_rate": 5.3769431716926456e-05, "loss": 0.026, "step": 80350 }, { "epoch": 1.6058510850885255, "grad_norm": 0.08161801844835281, "learning_rate": 5.3759126858869056e-05, "loss": 0.024, "step": 80360 }, { "epoch": 1.6060509172295272, "grad_norm": 0.13716574013233185, "learning_rate": 5.374882184022866e-05, "loss": 0.0257, "step": 80370 }, { "epoch": 1.606250749370529, "grad_norm": 0.10504662990570068, "learning_rate": 5.373851666144548e-05, "loss": 0.0287, "step": 80380 }, { "epoch": 1.6064505815115302, "grad_norm": 0.1157526895403862, "learning_rate": 5.372821132295975e-05, "loss": 0.0295, "step": 80390 }, { "epoch": 1.606650413652532, "grad_norm": 0.12033996731042862, "learning_rate": 5.371790582521167e-05, "loss": 0.0269, "step": 80400 }, { "epoch": 1.6068502457935334, "grad_norm": 0.13631165027618408, "learning_rate": 5.370760016864149e-05, "loss": 0.0324, "step": 80410 }, { "epoch": 1.607050077934535, "grad_norm": 0.13226278126239777, "learning_rate": 5.369729435368943e-05, "loss": 0.0313, "step": 80420 }, { "epoch": 1.6072499100755366, "grad_norm": 0.10406699031591415, "learning_rate": 5.3686988380795754e-05, "loss": 0.0255, "step": 80430 }, { "epoch": 1.6074497422165381, "grad_norm": 0.11710327118635178, "learning_rate": 5.3676682250400725e-05, "loss": 0.0323, "step": 80440 }, { "epoch": 1.6076495743575396, "grad_norm": 0.15622346103191376, "learning_rate": 5.366637596294455e-05, "loss": 0.0277, "step": 80450 }, { "epoch": 1.6078494064985414, "grad_norm": 0.13475805521011353, "learning_rate": 5.365606951886756e-05, "loss": 0.028, "step": 80460 }, { "epoch": 1.6080492386395426, "grad_norm": 0.21058984100818634, "learning_rate": 5.3645762918609986e-05, "loss": 0.0348, "step": 80470 }, { "epoch": 1.6082490707805444, "grad_norm": 0.13473373651504517, "learning_rate": 5.363545616261211e-05, "loss": 0.028, "step": 80480 }, { "epoch": 1.6084489029215459, "grad_norm": 0.14386679232120514, "learning_rate": 5.362514925131423e-05, "loss": 0.0356, "step": 80490 }, { "epoch": 1.6086487350625474, "grad_norm": 0.11143585294485092, "learning_rate": 5.3614842185156634e-05, "loss": 0.0261, "step": 80500 }, { "epoch": 1.608848567203549, "grad_norm": 0.13445191085338593, "learning_rate": 5.3604534964579624e-05, "loss": 0.0348, "step": 80510 }, { "epoch": 1.6090483993445506, "grad_norm": 0.14317750930786133, "learning_rate": 5.3594227590023496e-05, "loss": 0.0284, "step": 80520 }, { "epoch": 1.609248231485552, "grad_norm": 0.1023300439119339, "learning_rate": 5.358392006192857e-05, "loss": 0.0335, "step": 80530 }, { "epoch": 1.6094480636265538, "grad_norm": 0.13241207599639893, "learning_rate": 5.3573612380735175e-05, "loss": 0.0268, "step": 80540 }, { "epoch": 1.6096478957675553, "grad_norm": 0.13501259684562683, "learning_rate": 5.356330454688361e-05, "loss": 0.0297, "step": 80550 }, { "epoch": 1.6098477279085568, "grad_norm": 0.141554057598114, "learning_rate": 5.3552996560814215e-05, "loss": 0.0304, "step": 80560 }, { "epoch": 1.6100475600495585, "grad_norm": 0.11745588481426239, "learning_rate": 5.3542688422967344e-05, "loss": 0.0261, "step": 80570 }, { "epoch": 1.6102473921905598, "grad_norm": 0.12038059532642365, "learning_rate": 5.353238013378332e-05, "loss": 0.025, "step": 80580 }, { "epoch": 1.6104472243315615, "grad_norm": 0.14341284334659576, "learning_rate": 5.3522071693702536e-05, "loss": 0.0356, "step": 80590 }, { "epoch": 1.610647056472563, "grad_norm": 0.12502847611904144, "learning_rate": 5.351176310316528e-05, "loss": 0.0234, "step": 80600 }, { "epoch": 1.6108468886135645, "grad_norm": 0.1437007188796997, "learning_rate": 5.350145436261199e-05, "loss": 0.0353, "step": 80610 }, { "epoch": 1.6110467207545662, "grad_norm": 0.10087674111127853, "learning_rate": 5.3491145472482984e-05, "loss": 0.0313, "step": 80620 }, { "epoch": 1.6112465528955677, "grad_norm": 0.1366768330335617, "learning_rate": 5.348083643321866e-05, "loss": 0.0296, "step": 80630 }, { "epoch": 1.6114463850365692, "grad_norm": 0.13033604621887207, "learning_rate": 5.34705272452594e-05, "loss": 0.0359, "step": 80640 }, { "epoch": 1.611646217177571, "grad_norm": 0.0904327854514122, "learning_rate": 5.3460217909045585e-05, "loss": 0.0274, "step": 80650 }, { "epoch": 1.6118460493185724, "grad_norm": 0.14203067123889923, "learning_rate": 5.3449908425017626e-05, "loss": 0.0285, "step": 80660 }, { "epoch": 1.612045881459574, "grad_norm": 0.13712258636951447, "learning_rate": 5.3439598793615916e-05, "loss": 0.0419, "step": 80670 }, { "epoch": 1.6122457136005757, "grad_norm": 0.1458742320537567, "learning_rate": 5.342928901528086e-05, "loss": 0.0256, "step": 80680 }, { "epoch": 1.612445545741577, "grad_norm": 0.11341413110494614, "learning_rate": 5.341897909045288e-05, "loss": 0.0347, "step": 80690 }, { "epoch": 1.6126453778825787, "grad_norm": 0.11166355013847351, "learning_rate": 5.340866901957239e-05, "loss": 0.0289, "step": 80700 }, { "epoch": 1.6128452100235802, "grad_norm": 0.1246013268828392, "learning_rate": 5.339835880307982e-05, "loss": 0.0253, "step": 80710 }, { "epoch": 1.6130450421645817, "grad_norm": 0.1332600712776184, "learning_rate": 5.3388048441415604e-05, "loss": 0.0305, "step": 80720 }, { "epoch": 1.6132448743055834, "grad_norm": 0.15275143086910248, "learning_rate": 5.33777379350202e-05, "loss": 0.034, "step": 80730 }, { "epoch": 1.6134447064465849, "grad_norm": 0.14338618516921997, "learning_rate": 5.3367427284334026e-05, "loss": 0.0268, "step": 80740 }, { "epoch": 1.6136445385875864, "grad_norm": 0.11465633660554886, "learning_rate": 5.335711648979753e-05, "loss": 0.0329, "step": 80750 }, { "epoch": 1.613844370728588, "grad_norm": 0.1602051854133606, "learning_rate": 5.33468055518512e-05, "loss": 0.0264, "step": 80760 }, { "epoch": 1.6140442028695894, "grad_norm": 0.14926418662071228, "learning_rate": 5.333649447093548e-05, "loss": 0.0284, "step": 80770 }, { "epoch": 1.614244035010591, "grad_norm": 0.1261741816997528, "learning_rate": 5.332618324749085e-05, "loss": 0.0286, "step": 80780 }, { "epoch": 1.6144438671515928, "grad_norm": 0.1157030537724495, "learning_rate": 5.331587188195779e-05, "loss": 0.0318, "step": 80790 }, { "epoch": 1.614643699292594, "grad_norm": 0.17411178350448608, "learning_rate": 5.330556037477675e-05, "loss": 0.024, "step": 80800 }, { "epoch": 1.6148435314335958, "grad_norm": 0.1590064913034439, "learning_rate": 5.329524872638827e-05, "loss": 0.0309, "step": 80810 }, { "epoch": 1.6150433635745973, "grad_norm": 0.12567773461341858, "learning_rate": 5.32849369372328e-05, "loss": 0.0344, "step": 80820 }, { "epoch": 1.6152431957155988, "grad_norm": 0.11691281199455261, "learning_rate": 5.327462500775087e-05, "loss": 0.0378, "step": 80830 }, { "epoch": 1.6154430278566005, "grad_norm": 0.12169568240642548, "learning_rate": 5.3264312938382974e-05, "loss": 0.0379, "step": 80840 }, { "epoch": 1.615642859997602, "grad_norm": 0.16434724628925323, "learning_rate": 5.325400072956961e-05, "loss": 0.0369, "step": 80850 }, { "epoch": 1.6158426921386035, "grad_norm": 0.16603495180606842, "learning_rate": 5.324368838175133e-05, "loss": 0.0266, "step": 80860 }, { "epoch": 1.6160425242796053, "grad_norm": 0.0991894081234932, "learning_rate": 5.323337589536864e-05, "loss": 0.0369, "step": 80870 }, { "epoch": 1.6162423564206065, "grad_norm": 0.13864006102085114, "learning_rate": 5.322306327086206e-05, "loss": 0.0374, "step": 80880 }, { "epoch": 1.6164421885616083, "grad_norm": 0.1412583440542221, "learning_rate": 5.321275050867215e-05, "loss": 0.027, "step": 80890 }, { "epoch": 1.6166420207026098, "grad_norm": 0.15426458418369293, "learning_rate": 5.320243760923943e-05, "loss": 0.0307, "step": 80900 }, { "epoch": 1.6168418528436113, "grad_norm": 0.14142809808254242, "learning_rate": 5.3192124573004474e-05, "loss": 0.0324, "step": 80910 }, { "epoch": 1.617041684984613, "grad_norm": 0.1329142451286316, "learning_rate": 5.31818114004078e-05, "loss": 0.0351, "step": 80920 }, { "epoch": 1.6172415171256145, "grad_norm": 0.10342016816139221, "learning_rate": 5.317149809189e-05, "loss": 0.0302, "step": 80930 }, { "epoch": 1.617441349266616, "grad_norm": 0.11755421757698059, "learning_rate": 5.316118464789164e-05, "loss": 0.0255, "step": 80940 }, { "epoch": 1.6176411814076177, "grad_norm": 0.08452005684375763, "learning_rate": 5.3150871068853255e-05, "loss": 0.0368, "step": 80950 }, { "epoch": 1.6178410135486192, "grad_norm": 0.15022540092468262, "learning_rate": 5.314055735521546e-05, "loss": 0.0306, "step": 80960 }, { "epoch": 1.6180408456896207, "grad_norm": 0.1066068708896637, "learning_rate": 5.313024350741882e-05, "loss": 0.028, "step": 80970 }, { "epoch": 1.6182406778306224, "grad_norm": 0.10537995398044586, "learning_rate": 5.3119929525903924e-05, "loss": 0.0278, "step": 80980 }, { "epoch": 1.6184405099716237, "grad_norm": 0.11083099991083145, "learning_rate": 5.310961541111138e-05, "loss": 0.0293, "step": 80990 }, { "epoch": 1.6186403421126254, "grad_norm": 0.11247164011001587, "learning_rate": 5.3099301163481775e-05, "loss": 0.0285, "step": 81000 }, { "epoch": 1.618840174253627, "grad_norm": 0.12347281724214554, "learning_rate": 5.308898678345572e-05, "loss": 0.0344, "step": 81010 }, { "epoch": 1.6190400063946284, "grad_norm": 0.10457771271467209, "learning_rate": 5.3078672271473816e-05, "loss": 0.0284, "step": 81020 }, { "epoch": 1.6192398385356301, "grad_norm": 0.09007873386144638, "learning_rate": 5.306835762797669e-05, "loss": 0.0327, "step": 81030 }, { "epoch": 1.6194396706766316, "grad_norm": 0.11286448687314987, "learning_rate": 5.3058042853404974e-05, "loss": 0.029, "step": 81040 }, { "epoch": 1.6196395028176331, "grad_norm": 0.1250620186328888, "learning_rate": 5.304772794819927e-05, "loss": 0.03, "step": 81050 }, { "epoch": 1.6198393349586349, "grad_norm": 0.13778291642665863, "learning_rate": 5.303741291280023e-05, "loss": 0.0315, "step": 81060 }, { "epoch": 1.6200391670996364, "grad_norm": 0.1143423393368721, "learning_rate": 5.30270977476485e-05, "loss": 0.0251, "step": 81070 }, { "epoch": 1.6202389992406379, "grad_norm": 0.09654907882213593, "learning_rate": 5.301678245318471e-05, "loss": 0.0307, "step": 81080 }, { "epoch": 1.6204388313816396, "grad_norm": 0.11750580370426178, "learning_rate": 5.3006467029849513e-05, "loss": 0.0322, "step": 81090 }, { "epoch": 1.6206386635226409, "grad_norm": 0.11549991369247437, "learning_rate": 5.299615147808357e-05, "loss": 0.0231, "step": 81100 }, { "epoch": 1.6208384956636426, "grad_norm": 0.10137273371219635, "learning_rate": 5.298583579832754e-05, "loss": 0.0301, "step": 81110 }, { "epoch": 1.621038327804644, "grad_norm": 0.11594253033399582, "learning_rate": 5.2975519991022084e-05, "loss": 0.0301, "step": 81120 }, { "epoch": 1.6212381599456456, "grad_norm": 0.10242393612861633, "learning_rate": 5.2965204056607886e-05, "loss": 0.0276, "step": 81130 }, { "epoch": 1.6214379920866473, "grad_norm": 0.09972093254327774, "learning_rate": 5.295488799552562e-05, "loss": 0.0261, "step": 81140 }, { "epoch": 1.6216378242276488, "grad_norm": 0.1511009931564331, "learning_rate": 5.294457180821596e-05, "loss": 0.031, "step": 81150 }, { "epoch": 1.6218376563686503, "grad_norm": 0.1120658740401268, "learning_rate": 5.2934255495119614e-05, "loss": 0.0322, "step": 81160 }, { "epoch": 1.622037488509652, "grad_norm": 0.1376328468322754, "learning_rate": 5.292393905667724e-05, "loss": 0.0318, "step": 81170 }, { "epoch": 1.6222373206506533, "grad_norm": 0.11842738837003708, "learning_rate": 5.2913622493329594e-05, "loss": 0.0341, "step": 81180 }, { "epoch": 1.622437152791655, "grad_norm": 0.1453714668750763, "learning_rate": 5.290330580551732e-05, "loss": 0.0361, "step": 81190 }, { "epoch": 1.6226369849326567, "grad_norm": 0.11816854029893875, "learning_rate": 5.289298899368117e-05, "loss": 0.0263, "step": 81200 }, { "epoch": 1.622836817073658, "grad_norm": 0.13398180902004242, "learning_rate": 5.288267205826184e-05, "loss": 0.023, "step": 81210 }, { "epoch": 1.6230366492146597, "grad_norm": 0.16683733463287354, "learning_rate": 5.287235499970006e-05, "loss": 0.0299, "step": 81220 }, { "epoch": 1.6232364813556612, "grad_norm": 0.11071082949638367, "learning_rate": 5.2862037818436537e-05, "loss": 0.0409, "step": 81230 }, { "epoch": 1.6234363134966627, "grad_norm": 0.11320408433675766, "learning_rate": 5.285172051491204e-05, "loss": 0.0321, "step": 81240 }, { "epoch": 1.6236361456376645, "grad_norm": 0.147637277841568, "learning_rate": 5.2841403089567256e-05, "loss": 0.0291, "step": 81250 }, { "epoch": 1.623835977778666, "grad_norm": 0.12837518751621246, "learning_rate": 5.2831085542842974e-05, "loss": 0.0273, "step": 81260 }, { "epoch": 1.6240358099196675, "grad_norm": 0.12220212817192078, "learning_rate": 5.28207678751799e-05, "loss": 0.0227, "step": 81270 }, { "epoch": 1.6242356420606692, "grad_norm": 0.1315704733133316, "learning_rate": 5.2810450087018814e-05, "loss": 0.0258, "step": 81280 }, { "epoch": 1.6244354742016704, "grad_norm": 0.11740152537822723, "learning_rate": 5.2800132178800466e-05, "loss": 0.0357, "step": 81290 }, { "epoch": 1.6246353063426722, "grad_norm": 0.10016366839408875, "learning_rate": 5.27898141509656e-05, "loss": 0.031, "step": 81300 }, { "epoch": 1.6248351384836737, "grad_norm": 0.17385384440422058, "learning_rate": 5.277949600395502e-05, "loss": 0.0234, "step": 81310 }, { "epoch": 1.6250349706246752, "grad_norm": 0.13925834000110626, "learning_rate": 5.2769177738209454e-05, "loss": 0.0316, "step": 81320 }, { "epoch": 1.625234802765677, "grad_norm": 0.13594715297222137, "learning_rate": 5.275885935416972e-05, "loss": 0.0299, "step": 81330 }, { "epoch": 1.6254346349066784, "grad_norm": 0.11897746473550797, "learning_rate": 5.274854085227657e-05, "loss": 0.0282, "step": 81340 }, { "epoch": 1.6256344670476799, "grad_norm": 0.13043883442878723, "learning_rate": 5.273822223297081e-05, "loss": 0.0297, "step": 81350 }, { "epoch": 1.6258342991886816, "grad_norm": 0.13341715931892395, "learning_rate": 5.272790349669323e-05, "loss": 0.0279, "step": 81360 }, { "epoch": 1.626034131329683, "grad_norm": 0.16622917354106903, "learning_rate": 5.271758464388462e-05, "loss": 0.0296, "step": 81370 }, { "epoch": 1.6262339634706846, "grad_norm": 0.16564711928367615, "learning_rate": 5.27072656749858e-05, "loss": 0.0305, "step": 81380 }, { "epoch": 1.6264337956116863, "grad_norm": 0.15301960706710815, "learning_rate": 5.269694659043754e-05, "loss": 0.0329, "step": 81390 }, { "epoch": 1.6266336277526876, "grad_norm": 0.08914031833410263, "learning_rate": 5.2686627390680684e-05, "loss": 0.0254, "step": 81400 }, { "epoch": 1.6268334598936893, "grad_norm": 0.12118427455425262, "learning_rate": 5.2676308076156046e-05, "loss": 0.0281, "step": 81410 }, { "epoch": 1.6270332920346908, "grad_norm": 0.1428389996290207, "learning_rate": 5.266598864730444e-05, "loss": 0.0279, "step": 81420 }, { "epoch": 1.6272331241756923, "grad_norm": 0.13424448668956757, "learning_rate": 5.26556691045667e-05, "loss": 0.0388, "step": 81430 }, { "epoch": 1.627432956316694, "grad_norm": 0.13578329980373383, "learning_rate": 5.264534944838366e-05, "loss": 0.0326, "step": 81440 }, { "epoch": 1.6276327884576955, "grad_norm": 0.12785251438617706, "learning_rate": 5.263502967919613e-05, "loss": 0.0247, "step": 81450 }, { "epoch": 1.627832620598697, "grad_norm": 0.16473424434661865, "learning_rate": 5.2624709797445005e-05, "loss": 0.0344, "step": 81460 }, { "epoch": 1.6280324527396988, "grad_norm": 0.14053966104984283, "learning_rate": 5.2614389803571076e-05, "loss": 0.0321, "step": 81470 }, { "epoch": 1.6282322848807003, "grad_norm": 0.1536126434803009, "learning_rate": 5.260406969801522e-05, "loss": 0.0337, "step": 81480 }, { "epoch": 1.6284321170217018, "grad_norm": 0.14043821394443512, "learning_rate": 5.25937494812183e-05, "loss": 0.0355, "step": 81490 }, { "epoch": 1.6286319491627035, "grad_norm": 0.1088043749332428, "learning_rate": 5.2583429153621156e-05, "loss": 0.0378, "step": 81500 }, { "epoch": 1.6288317813037048, "grad_norm": 0.14380058646202087, "learning_rate": 5.257310871566468e-05, "loss": 0.0295, "step": 81510 }, { "epoch": 1.6290316134447065, "grad_norm": 0.1155250072479248, "learning_rate": 5.25627881677897e-05, "loss": 0.0383, "step": 81520 }, { "epoch": 1.629231445585708, "grad_norm": 0.1443967968225479, "learning_rate": 5.255246751043714e-05, "loss": 0.0226, "step": 81530 }, { "epoch": 1.6294312777267095, "grad_norm": 0.09763778746128082, "learning_rate": 5.254214674404785e-05, "loss": 0.0209, "step": 81540 }, { "epoch": 1.6296311098677112, "grad_norm": 0.12679219245910645, "learning_rate": 5.2531825869062715e-05, "loss": 0.032, "step": 81550 }, { "epoch": 1.6298309420087127, "grad_norm": 0.12969644367694855, "learning_rate": 5.252150488592265e-05, "loss": 0.0302, "step": 81560 }, { "epoch": 1.6300307741497142, "grad_norm": 0.11254438757896423, "learning_rate": 5.25111837950685e-05, "loss": 0.0408, "step": 81570 }, { "epoch": 1.630230606290716, "grad_norm": 0.13744640350341797, "learning_rate": 5.250086259694121e-05, "loss": 0.0329, "step": 81580 }, { "epoch": 1.6304304384317172, "grad_norm": 0.13334117829799652, "learning_rate": 5.249054129198166e-05, "loss": 0.0241, "step": 81590 }, { "epoch": 1.630630270572719, "grad_norm": 0.12845765054225922, "learning_rate": 5.2480219880630765e-05, "loss": 0.0277, "step": 81600 }, { "epoch": 1.6308301027137204, "grad_norm": 0.1593729853630066, "learning_rate": 5.246989836332943e-05, "loss": 0.0306, "step": 81610 }, { "epoch": 1.631029934854722, "grad_norm": 0.12256614118814468, "learning_rate": 5.245957674051857e-05, "loss": 0.0309, "step": 81620 }, { "epoch": 1.6312297669957236, "grad_norm": 0.1147068440914154, "learning_rate": 5.244925501263911e-05, "loss": 0.0275, "step": 81630 }, { "epoch": 1.6314295991367251, "grad_norm": 0.12870082259178162, "learning_rate": 5.2438933180131986e-05, "loss": 0.021, "step": 81640 }, { "epoch": 1.6316294312777266, "grad_norm": 0.13041618466377258, "learning_rate": 5.242861124343811e-05, "loss": 0.0236, "step": 81650 }, { "epoch": 1.6318292634187284, "grad_norm": 0.12968359887599945, "learning_rate": 5.2418289202998414e-05, "loss": 0.0297, "step": 81660 }, { "epoch": 1.6320290955597299, "grad_norm": 0.12800703942775726, "learning_rate": 5.240796705925386e-05, "loss": 0.0273, "step": 81670 }, { "epoch": 1.6322289277007314, "grad_norm": 0.12007931619882584, "learning_rate": 5.239764481264536e-05, "loss": 0.0292, "step": 81680 }, { "epoch": 1.632428759841733, "grad_norm": 0.09921383112668991, "learning_rate": 5.238732246361391e-05, "loss": 0.029, "step": 81690 }, { "epoch": 1.6326285919827344, "grad_norm": 0.12498138099908829, "learning_rate": 5.23770000126004e-05, "loss": 0.0301, "step": 81700 }, { "epoch": 1.632828424123736, "grad_norm": 0.11467217653989792, "learning_rate": 5.236667746004583e-05, "loss": 0.0278, "step": 81710 }, { "epoch": 1.6330282562647376, "grad_norm": 0.11010532081127167, "learning_rate": 5.235635480639113e-05, "loss": 0.0283, "step": 81720 }, { "epoch": 1.633228088405739, "grad_norm": 0.10698923468589783, "learning_rate": 5.23460320520773e-05, "loss": 0.0271, "step": 81730 }, { "epoch": 1.6334279205467408, "grad_norm": 0.13918812572956085, "learning_rate": 5.233570919754528e-05, "loss": 0.0352, "step": 81740 }, { "epoch": 1.6336277526877423, "grad_norm": 0.11759039759635925, "learning_rate": 5.232538624323606e-05, "loss": 0.0253, "step": 81750 }, { "epoch": 1.6338275848287438, "grad_norm": 0.15513694286346436, "learning_rate": 5.231506318959062e-05, "loss": 0.0321, "step": 81760 }, { "epoch": 1.6340274169697455, "grad_norm": 0.11260222643613815, "learning_rate": 5.2304740037049916e-05, "loss": 0.0289, "step": 81770 }, { "epoch": 1.634227249110747, "grad_norm": 0.16063129901885986, "learning_rate": 5.2294416786054966e-05, "loss": 0.0313, "step": 81780 }, { "epoch": 1.6344270812517485, "grad_norm": 0.11548474431037903, "learning_rate": 5.228409343704673e-05, "loss": 0.0273, "step": 81790 }, { "epoch": 1.6346269133927502, "grad_norm": 0.12208515405654907, "learning_rate": 5.227376999046622e-05, "loss": 0.0309, "step": 81800 }, { "epoch": 1.6348267455337515, "grad_norm": 0.1458144187927246, "learning_rate": 5.226344644675445e-05, "loss": 0.0278, "step": 81810 }, { "epoch": 1.6350265776747532, "grad_norm": 0.16329099237918854, "learning_rate": 5.22531228063524e-05, "loss": 0.0306, "step": 81820 }, { "epoch": 1.6352264098157547, "grad_norm": 0.13651329278945923, "learning_rate": 5.2242799069701065e-05, "loss": 0.0281, "step": 81830 }, { "epoch": 1.6354262419567562, "grad_norm": 0.11739542335271835, "learning_rate": 5.2232475237241504e-05, "loss": 0.0222, "step": 81840 }, { "epoch": 1.635626074097758, "grad_norm": 0.13331864774227142, "learning_rate": 5.222215130941468e-05, "loss": 0.03, "step": 81850 }, { "epoch": 1.6358259062387595, "grad_norm": 0.11133521050214767, "learning_rate": 5.221182728666164e-05, "loss": 0.0316, "step": 81860 }, { "epoch": 1.636025738379761, "grad_norm": 0.13301324844360352, "learning_rate": 5.220150316942341e-05, "loss": 0.0233, "step": 81870 }, { "epoch": 1.6362255705207627, "grad_norm": 0.1558436006307602, "learning_rate": 5.2191178958140986e-05, "loss": 0.0321, "step": 81880 }, { "epoch": 1.6364254026617642, "grad_norm": 0.157695472240448, "learning_rate": 5.218085465325545e-05, "loss": 0.0339, "step": 81890 }, { "epoch": 1.6366252348027657, "grad_norm": 0.14358191192150116, "learning_rate": 5.217053025520778e-05, "loss": 0.0331, "step": 81900 }, { "epoch": 1.6368250669437674, "grad_norm": 0.11193543672561646, "learning_rate": 5.216020576443907e-05, "loss": 0.0304, "step": 81910 }, { "epoch": 1.6370248990847687, "grad_norm": 0.1237606331706047, "learning_rate": 5.2149881181390325e-05, "loss": 0.0424, "step": 81920 }, { "epoch": 1.6372247312257704, "grad_norm": 0.11508557200431824, "learning_rate": 5.213955650650262e-05, "loss": 0.0284, "step": 81930 }, { "epoch": 1.637424563366772, "grad_norm": 0.11424627155065536, "learning_rate": 5.212923174021698e-05, "loss": 0.0311, "step": 81940 }, { "epoch": 1.6376243955077734, "grad_norm": 0.13285411894321442, "learning_rate": 5.211890688297447e-05, "loss": 0.0299, "step": 81950 }, { "epoch": 1.637824227648775, "grad_norm": 0.1399119794368744, "learning_rate": 5.2108581935216164e-05, "loss": 0.0357, "step": 81960 }, { "epoch": 1.6380240597897766, "grad_norm": 0.18188032507896423, "learning_rate": 5.209825689738308e-05, "loss": 0.0259, "step": 81970 }, { "epoch": 1.638223891930778, "grad_norm": 0.11807454377412796, "learning_rate": 5.2087931769916355e-05, "loss": 0.0249, "step": 81980 }, { "epoch": 1.6384237240717798, "grad_norm": 0.1262485384941101, "learning_rate": 5.2077606553257e-05, "loss": 0.0284, "step": 81990 }, { "epoch": 1.638623556212781, "grad_norm": 0.13671866059303284, "learning_rate": 5.206728124784611e-05, "loss": 0.0248, "step": 82000 }, { "epoch": 1.6388233883537828, "grad_norm": 0.12638042867183685, "learning_rate": 5.205695585412476e-05, "loss": 0.03, "step": 82010 }, { "epoch": 1.6390232204947843, "grad_norm": 0.13427357375621796, "learning_rate": 5.2046630372534034e-05, "loss": 0.0298, "step": 82020 }, { "epoch": 1.6392230526357858, "grad_norm": 0.08277347683906555, "learning_rate": 5.2036304803515037e-05, "loss": 0.0278, "step": 82030 }, { "epoch": 1.6394228847767875, "grad_norm": 0.11395802348852158, "learning_rate": 5.202597914750883e-05, "loss": 0.0307, "step": 82040 }, { "epoch": 1.639622716917789, "grad_norm": 0.1323985904455185, "learning_rate": 5.2015653404956513e-05, "loss": 0.0275, "step": 82050 }, { "epoch": 1.6398225490587905, "grad_norm": 0.09484227746725082, "learning_rate": 5.2005327576299176e-05, "loss": 0.0331, "step": 82060 }, { "epoch": 1.6400223811997923, "grad_norm": 0.11022281646728516, "learning_rate": 5.1995001661977934e-05, "loss": 0.0252, "step": 82070 }, { "epoch": 1.6402222133407938, "grad_norm": 0.10723742842674255, "learning_rate": 5.198467566243389e-05, "loss": 0.0238, "step": 82080 }, { "epoch": 1.6404220454817953, "grad_norm": 0.1446230709552765, "learning_rate": 5.197434957810815e-05, "loss": 0.0336, "step": 82090 }, { "epoch": 1.640621877622797, "grad_norm": 0.12430528551340103, "learning_rate": 5.196402340944181e-05, "loss": 0.0273, "step": 82100 }, { "epoch": 1.6408217097637983, "grad_norm": 0.1507774293422699, "learning_rate": 5.195369715687603e-05, "loss": 0.0342, "step": 82110 }, { "epoch": 1.6410215419048, "grad_norm": 0.16409049928188324, "learning_rate": 5.194337082085187e-05, "loss": 0.0287, "step": 82120 }, { "epoch": 1.6412213740458015, "grad_norm": 0.09349633008241653, "learning_rate": 5.1933044401810474e-05, "loss": 0.0299, "step": 82130 }, { "epoch": 1.641421206186803, "grad_norm": 0.11584822833538055, "learning_rate": 5.1922717900192976e-05, "loss": 0.0365, "step": 82140 }, { "epoch": 1.6416210383278047, "grad_norm": 0.13380049169063568, "learning_rate": 5.1912391316440515e-05, "loss": 0.0255, "step": 82150 }, { "epoch": 1.6418208704688062, "grad_norm": 0.17856228351593018, "learning_rate": 5.1902064650994206e-05, "loss": 0.0354, "step": 82160 }, { "epoch": 1.6420207026098077, "grad_norm": 0.1066674217581749, "learning_rate": 5.189173790429517e-05, "loss": 0.0273, "step": 82170 }, { "epoch": 1.6422205347508094, "grad_norm": 0.1329800933599472, "learning_rate": 5.188141107678459e-05, "loss": 0.0288, "step": 82180 }, { "epoch": 1.642420366891811, "grad_norm": 0.11989545077085495, "learning_rate": 5.187108416890357e-05, "loss": 0.0295, "step": 82190 }, { "epoch": 1.6426201990328124, "grad_norm": 0.14033153653144836, "learning_rate": 5.186075718109328e-05, "loss": 0.0281, "step": 82200 }, { "epoch": 1.6428200311738141, "grad_norm": 0.1213831901550293, "learning_rate": 5.185043011379487e-05, "loss": 0.0251, "step": 82210 }, { "epoch": 1.6430198633148154, "grad_norm": 0.16714400053024292, "learning_rate": 5.1840102967449466e-05, "loss": 0.0313, "step": 82220 }, { "epoch": 1.6432196954558171, "grad_norm": 0.09045758843421936, "learning_rate": 5.1829775742498254e-05, "loss": 0.0297, "step": 82230 }, { "epoch": 1.6434195275968186, "grad_norm": 0.11086783558130264, "learning_rate": 5.181944843938238e-05, "loss": 0.028, "step": 82240 }, { "epoch": 1.6436193597378201, "grad_norm": 0.12724639475345612, "learning_rate": 5.180912105854301e-05, "loss": 0.0303, "step": 82250 }, { "epoch": 1.6438191918788219, "grad_norm": 0.09052006900310516, "learning_rate": 5.1798793600421314e-05, "loss": 0.032, "step": 82260 }, { "epoch": 1.6440190240198234, "grad_norm": 0.08861726522445679, "learning_rate": 5.178846606545846e-05, "loss": 0.0228, "step": 82270 }, { "epoch": 1.6442188561608249, "grad_norm": 0.136774942278862, "learning_rate": 5.1778138454095626e-05, "loss": 0.0342, "step": 82280 }, { "epoch": 1.6444186883018266, "grad_norm": 0.12107329815626144, "learning_rate": 5.1767810766773984e-05, "loss": 0.0346, "step": 82290 }, { "epoch": 1.6446185204428279, "grad_norm": 0.14536064863204956, "learning_rate": 5.1757483003934706e-05, "loss": 0.0292, "step": 82300 }, { "epoch": 1.6448183525838296, "grad_norm": 0.21845290064811707, "learning_rate": 5.174715516601899e-05, "loss": 0.0345, "step": 82310 }, { "epoch": 1.6450181847248313, "grad_norm": 0.17313922941684723, "learning_rate": 5.1736827253468e-05, "loss": 0.0333, "step": 82320 }, { "epoch": 1.6452180168658326, "grad_norm": 0.17803210020065308, "learning_rate": 5.172649926672295e-05, "loss": 0.0302, "step": 82330 }, { "epoch": 1.6454178490068343, "grad_norm": 0.14908722043037415, "learning_rate": 5.171617120622503e-05, "loss": 0.0331, "step": 82340 }, { "epoch": 1.6456176811478358, "grad_norm": 0.16869980096817017, "learning_rate": 5.1705843072415415e-05, "loss": 0.0312, "step": 82350 }, { "epoch": 1.6458175132888373, "grad_norm": 0.13750059902668, "learning_rate": 5.169551486573534e-05, "loss": 0.0222, "step": 82360 }, { "epoch": 1.646017345429839, "grad_norm": 0.1583654284477234, "learning_rate": 5.1685186586625955e-05, "loss": 0.0309, "step": 82370 }, { "epoch": 1.6462171775708405, "grad_norm": 0.15451575815677643, "learning_rate": 5.167485823552852e-05, "loss": 0.0264, "step": 82380 }, { "epoch": 1.646417009711842, "grad_norm": 0.1269569844007492, "learning_rate": 5.166452981288421e-05, "loss": 0.0385, "step": 82390 }, { "epoch": 1.6466168418528437, "grad_norm": 0.1584596186876297, "learning_rate": 5.165420131913423e-05, "loss": 0.0302, "step": 82400 }, { "epoch": 1.646816673993845, "grad_norm": 0.13872481882572174, "learning_rate": 5.164387275471983e-05, "loss": 0.0291, "step": 82410 }, { "epoch": 1.6470165061348467, "grad_norm": 0.1545099914073944, "learning_rate": 5.163354412008219e-05, "loss": 0.0296, "step": 82420 }, { "epoch": 1.6472163382758482, "grad_norm": 0.2293333113193512, "learning_rate": 5.1623215415662566e-05, "loss": 0.0299, "step": 82430 }, { "epoch": 1.6474161704168497, "grad_norm": 0.12132148444652557, "learning_rate": 5.161288664190214e-05, "loss": 0.0302, "step": 82440 }, { "epoch": 1.6476160025578515, "grad_norm": 0.12408372014760971, "learning_rate": 5.160255779924217e-05, "loss": 0.0282, "step": 82450 }, { "epoch": 1.647815834698853, "grad_norm": 0.16397510468959808, "learning_rate": 5.159222888812387e-05, "loss": 0.0253, "step": 82460 }, { "epoch": 1.6480156668398545, "grad_norm": 0.12451044470071793, "learning_rate": 5.158189990898847e-05, "loss": 0.0304, "step": 82470 }, { "epoch": 1.6482154989808562, "grad_norm": 0.15193438529968262, "learning_rate": 5.157157086227723e-05, "loss": 0.0252, "step": 82480 }, { "epoch": 1.6484153311218577, "grad_norm": 0.11327962577342987, "learning_rate": 5.156124174843137e-05, "loss": 0.0337, "step": 82490 }, { "epoch": 1.6486151632628592, "grad_norm": 0.1297367364168167, "learning_rate": 5.1550912567892115e-05, "loss": 0.0252, "step": 82500 }, { "epoch": 1.648814995403861, "grad_norm": 0.1619732826948166, "learning_rate": 5.154058332110073e-05, "loss": 0.0324, "step": 82510 }, { "epoch": 1.6490148275448622, "grad_norm": 0.13080920279026031, "learning_rate": 5.1530254008498455e-05, "loss": 0.0285, "step": 82520 }, { "epoch": 1.649214659685864, "grad_norm": 0.1332317590713501, "learning_rate": 5.151992463052654e-05, "loss": 0.0235, "step": 82530 }, { "epoch": 1.6494144918268654, "grad_norm": 0.12113150954246521, "learning_rate": 5.150959518762623e-05, "loss": 0.0283, "step": 82540 }, { "epoch": 1.649614323967867, "grad_norm": 0.11853020638227463, "learning_rate": 5.14992656802388e-05, "loss": 0.0344, "step": 82550 }, { "epoch": 1.6498141561088686, "grad_norm": 0.13745500147342682, "learning_rate": 5.1488936108805495e-05, "loss": 0.0269, "step": 82560 }, { "epoch": 1.6500139882498701, "grad_norm": 0.13487745821475983, "learning_rate": 5.1478606473767566e-05, "loss": 0.0297, "step": 82570 }, { "epoch": 1.6502138203908716, "grad_norm": 0.15485380589962006, "learning_rate": 5.1468276775566295e-05, "loss": 0.0231, "step": 82580 }, { "epoch": 1.6504136525318733, "grad_norm": 0.1695406436920166, "learning_rate": 5.1457947014642926e-05, "loss": 0.0279, "step": 82590 }, { "epoch": 1.6506134846728748, "grad_norm": 0.12447953969240189, "learning_rate": 5.144761719143875e-05, "loss": 0.0339, "step": 82600 }, { "epoch": 1.6508133168138763, "grad_norm": 0.12601643800735474, "learning_rate": 5.143728730639502e-05, "loss": 0.0238, "step": 82610 }, { "epoch": 1.651013148954878, "grad_norm": 0.14344024658203125, "learning_rate": 5.142695735995301e-05, "loss": 0.0342, "step": 82620 }, { "epoch": 1.6512129810958793, "grad_norm": 0.11966371536254883, "learning_rate": 5.141662735255402e-05, "loss": 0.0298, "step": 82630 }, { "epoch": 1.651412813236881, "grad_norm": 0.14372608065605164, "learning_rate": 5.140629728463929e-05, "loss": 0.0315, "step": 82640 }, { "epoch": 1.6516126453778825, "grad_norm": 0.14791704714298248, "learning_rate": 5.1395967156650135e-05, "loss": 0.0329, "step": 82650 }, { "epoch": 1.651812477518884, "grad_norm": 0.10976386070251465, "learning_rate": 5.1385636969027826e-05, "loss": 0.0275, "step": 82660 }, { "epoch": 1.6520123096598858, "grad_norm": 0.13157610595226288, "learning_rate": 5.1375306722213656e-05, "loss": 0.021, "step": 82670 }, { "epoch": 1.6522121418008873, "grad_norm": 0.09325981885194778, "learning_rate": 5.1364976416648915e-05, "loss": 0.0259, "step": 82680 }, { "epoch": 1.6524119739418888, "grad_norm": 0.13312415778636932, "learning_rate": 5.135464605277488e-05, "loss": 0.0295, "step": 82690 }, { "epoch": 1.6526118060828905, "grad_norm": 0.17062999308109283, "learning_rate": 5.134431563103285e-05, "loss": 0.0342, "step": 82700 }, { "epoch": 1.6528116382238918, "grad_norm": 0.11892391741275787, "learning_rate": 5.133398515186413e-05, "loss": 0.0325, "step": 82710 }, { "epoch": 1.6530114703648935, "grad_norm": 0.10258428007364273, "learning_rate": 5.132365461571002e-05, "loss": 0.0253, "step": 82720 }, { "epoch": 1.6532113025058952, "grad_norm": 0.15886929631233215, "learning_rate": 5.13133240230118e-05, "loss": 0.0307, "step": 82730 }, { "epoch": 1.6534111346468965, "grad_norm": 0.11151979118585587, "learning_rate": 5.130299337421081e-05, "loss": 0.0365, "step": 82740 }, { "epoch": 1.6536109667878982, "grad_norm": 0.1518823802471161, "learning_rate": 5.129266266974833e-05, "loss": 0.0297, "step": 82750 }, { "epoch": 1.6538107989288997, "grad_norm": 0.10667549073696136, "learning_rate": 5.128233191006567e-05, "loss": 0.0272, "step": 82760 }, { "epoch": 1.6540106310699012, "grad_norm": 0.11182893812656403, "learning_rate": 5.1272001095604164e-05, "loss": 0.0373, "step": 82770 }, { "epoch": 1.654210463210903, "grad_norm": 0.10636164247989655, "learning_rate": 5.126167022680508e-05, "loss": 0.0307, "step": 82780 }, { "epoch": 1.6544102953519044, "grad_norm": 0.13180765509605408, "learning_rate": 5.125133930410978e-05, "loss": 0.0251, "step": 82790 }, { "epoch": 1.654610127492906, "grad_norm": 0.12135082483291626, "learning_rate": 5.124100832795956e-05, "loss": 0.0264, "step": 82800 }, { "epoch": 1.6548099596339076, "grad_norm": 0.11012257635593414, "learning_rate": 5.1230677298795745e-05, "loss": 0.0317, "step": 82810 }, { "epoch": 1.655009791774909, "grad_norm": 0.09569552540779114, "learning_rate": 5.122034621705965e-05, "loss": 0.0325, "step": 82820 }, { "epoch": 1.6552096239159106, "grad_norm": 0.12794774770736694, "learning_rate": 5.121001508319262e-05, "loss": 0.0249, "step": 82830 }, { "epoch": 1.6554094560569121, "grad_norm": 0.15561850368976593, "learning_rate": 5.119968389763597e-05, "loss": 0.0265, "step": 82840 }, { "epoch": 1.6556092881979136, "grad_norm": 0.11420124024152756, "learning_rate": 5.1189352660831006e-05, "loss": 0.0366, "step": 82850 }, { "epoch": 1.6558091203389154, "grad_norm": 0.1363409161567688, "learning_rate": 5.1179021373219105e-05, "loss": 0.0366, "step": 82860 }, { "epoch": 1.6560089524799169, "grad_norm": 0.10534834116697311, "learning_rate": 5.1168690035241554e-05, "loss": 0.0279, "step": 82870 }, { "epoch": 1.6562087846209184, "grad_norm": 0.11659202724695206, "learning_rate": 5.115835864733973e-05, "loss": 0.0298, "step": 82880 }, { "epoch": 1.65640861676192, "grad_norm": 0.15954343974590302, "learning_rate": 5.114802720995493e-05, "loss": 0.0378, "step": 82890 }, { "epoch": 1.6566084489029216, "grad_norm": 0.14605392515659332, "learning_rate": 5.113769572352855e-05, "loss": 0.0271, "step": 82900 }, { "epoch": 1.656808281043923, "grad_norm": 0.13153037428855896, "learning_rate": 5.112736418850188e-05, "loss": 0.0274, "step": 82910 }, { "epoch": 1.6570081131849248, "grad_norm": 0.10245298594236374, "learning_rate": 5.1117032605316286e-05, "loss": 0.029, "step": 82920 }, { "epoch": 1.657207945325926, "grad_norm": 0.12193480134010315, "learning_rate": 5.11067009744131e-05, "loss": 0.029, "step": 82930 }, { "epoch": 1.6574077774669278, "grad_norm": 0.12061917781829834, "learning_rate": 5.109636929623369e-05, "loss": 0.0275, "step": 82940 }, { "epoch": 1.6576076096079293, "grad_norm": 0.12844477593898773, "learning_rate": 5.10860375712194e-05, "loss": 0.0312, "step": 82950 }, { "epoch": 1.6578074417489308, "grad_norm": 0.0965891033411026, "learning_rate": 5.107570579981158e-05, "loss": 0.0186, "step": 82960 }, { "epoch": 1.6580072738899325, "grad_norm": 0.12033107131719589, "learning_rate": 5.106537398245158e-05, "loss": 0.0336, "step": 82970 }, { "epoch": 1.658207106030934, "grad_norm": 0.12189027667045593, "learning_rate": 5.1055042119580764e-05, "loss": 0.0284, "step": 82980 }, { "epoch": 1.6584069381719355, "grad_norm": 0.18260417878627777, "learning_rate": 5.104471021164049e-05, "loss": 0.0263, "step": 82990 }, { "epoch": 1.6586067703129372, "grad_norm": 0.14323264360427856, "learning_rate": 5.1034378259072115e-05, "loss": 0.0306, "step": 83000 }, { "epoch": 1.6588066024539387, "grad_norm": 0.1667851209640503, "learning_rate": 5.102404626231701e-05, "loss": 0.0243, "step": 83010 }, { "epoch": 1.6590064345949402, "grad_norm": 0.1235933005809784, "learning_rate": 5.101371422181649e-05, "loss": 0.0288, "step": 83020 }, { "epoch": 1.659206266735942, "grad_norm": 0.11981939524412155, "learning_rate": 5.1003382138012e-05, "loss": 0.0272, "step": 83030 }, { "epoch": 1.6594060988769432, "grad_norm": 0.12193252891302109, "learning_rate": 5.0993050011344856e-05, "loss": 0.0239, "step": 83040 }, { "epoch": 1.659605931017945, "grad_norm": 0.10385442525148392, "learning_rate": 5.0982717842256436e-05, "loss": 0.0304, "step": 83050 }, { "epoch": 1.6598057631589465, "grad_norm": 0.13334456086158752, "learning_rate": 5.0972385631188135e-05, "loss": 0.0291, "step": 83060 }, { "epoch": 1.660005595299948, "grad_norm": 0.17950338125228882, "learning_rate": 5.096205337858128e-05, "loss": 0.0237, "step": 83070 }, { "epoch": 1.6602054274409497, "grad_norm": 0.15264354646205902, "learning_rate": 5.095172108487729e-05, "loss": 0.032, "step": 83080 }, { "epoch": 1.6604052595819512, "grad_norm": 0.17063598334789276, "learning_rate": 5.0941388750517494e-05, "loss": 0.0305, "step": 83090 }, { "epoch": 1.6606050917229527, "grad_norm": 0.1454668790102005, "learning_rate": 5.093105637594333e-05, "loss": 0.0327, "step": 83100 }, { "epoch": 1.6608049238639544, "grad_norm": 0.14869561791419983, "learning_rate": 5.092072396159614e-05, "loss": 0.0287, "step": 83110 }, { "epoch": 1.6610047560049557, "grad_norm": 0.07183009386062622, "learning_rate": 5.091039150791732e-05, "loss": 0.0244, "step": 83120 }, { "epoch": 1.6612045881459574, "grad_norm": 0.28398245573043823, "learning_rate": 5.0900059015348236e-05, "loss": 0.0305, "step": 83130 }, { "epoch": 1.6614044202869591, "grad_norm": 0.09619344025850296, "learning_rate": 5.0889726484330294e-05, "loss": 0.0291, "step": 83140 }, { "epoch": 1.6616042524279604, "grad_norm": 0.14994581043720245, "learning_rate": 5.0879393915304874e-05, "loss": 0.0224, "step": 83150 }, { "epoch": 1.6618040845689621, "grad_norm": 0.19805502891540527, "learning_rate": 5.086906130871335e-05, "loss": 0.027, "step": 83160 }, { "epoch": 1.6620039167099636, "grad_norm": 0.12141195684671402, "learning_rate": 5.085872866499712e-05, "loss": 0.019, "step": 83170 }, { "epoch": 1.6622037488509651, "grad_norm": 0.13744908571243286, "learning_rate": 5.084839598459759e-05, "loss": 0.0312, "step": 83180 }, { "epoch": 1.6624035809919668, "grad_norm": 0.09732603281736374, "learning_rate": 5.083806326795614e-05, "loss": 0.0299, "step": 83190 }, { "epoch": 1.6626034131329683, "grad_norm": 0.10906632989645004, "learning_rate": 5.082773051551416e-05, "loss": 0.0238, "step": 83200 }, { "epoch": 1.6628032452739698, "grad_norm": 0.12129650264978409, "learning_rate": 5.081739772771307e-05, "loss": 0.0253, "step": 83210 }, { "epoch": 1.6630030774149716, "grad_norm": 0.12450595945119858, "learning_rate": 5.080706490499422e-05, "loss": 0.033, "step": 83220 }, { "epoch": 1.6632029095559728, "grad_norm": 0.13422420620918274, "learning_rate": 5.079673204779907e-05, "loss": 0.0318, "step": 83230 }, { "epoch": 1.6634027416969746, "grad_norm": 0.12802429497241974, "learning_rate": 5.0786399156568974e-05, "loss": 0.0362, "step": 83240 }, { "epoch": 1.663602573837976, "grad_norm": 0.12864813208580017, "learning_rate": 5.077606623174536e-05, "loss": 0.0253, "step": 83250 }, { "epoch": 1.6638024059789775, "grad_norm": 0.13338670134544373, "learning_rate": 5.0765733273769625e-05, "loss": 0.0236, "step": 83260 }, { "epoch": 1.6640022381199793, "grad_norm": 0.11909334361553192, "learning_rate": 5.075540028308315e-05, "loss": 0.0376, "step": 83270 }, { "epoch": 1.6642020702609808, "grad_norm": 0.13028404116630554, "learning_rate": 5.074506726012739e-05, "loss": 0.0261, "step": 83280 }, { "epoch": 1.6644019024019823, "grad_norm": 0.1277729719877243, "learning_rate": 5.073473420534369e-05, "loss": 0.0371, "step": 83290 }, { "epoch": 1.664601734542984, "grad_norm": 0.11771663278341293, "learning_rate": 5.072440111917354e-05, "loss": 0.0266, "step": 83300 }, { "epoch": 1.6648015666839855, "grad_norm": 0.12687405943870544, "learning_rate": 5.0714068002058265e-05, "loss": 0.0361, "step": 83310 }, { "epoch": 1.665001398824987, "grad_norm": 0.17337261140346527, "learning_rate": 5.070373485443934e-05, "loss": 0.0249, "step": 83320 }, { "epoch": 1.6652012309659887, "grad_norm": 0.11159373819828033, "learning_rate": 5.069340167675816e-05, "loss": 0.025, "step": 83330 }, { "epoch": 1.66540106310699, "grad_norm": 0.15398260951042175, "learning_rate": 5.068306846945611e-05, "loss": 0.0298, "step": 83340 }, { "epoch": 1.6656008952479917, "grad_norm": 0.09589100629091263, "learning_rate": 5.0672735232974646e-05, "loss": 0.024, "step": 83350 }, { "epoch": 1.6658007273889932, "grad_norm": 0.10521911084651947, "learning_rate": 5.066240196775517e-05, "loss": 0.0252, "step": 83360 }, { "epoch": 1.6660005595299947, "grad_norm": 0.132421612739563, "learning_rate": 5.065206867423908e-05, "loss": 0.0372, "step": 83370 }, { "epoch": 1.6662003916709964, "grad_norm": 0.13558919727802277, "learning_rate": 5.0641735352867836e-05, "loss": 0.0327, "step": 83380 }, { "epoch": 1.666400223811998, "grad_norm": 0.14669018983840942, "learning_rate": 5.063140200408283e-05, "loss": 0.0231, "step": 83390 }, { "epoch": 1.6666000559529994, "grad_norm": 0.11645136773586273, "learning_rate": 5.0621068628325484e-05, "loss": 0.0267, "step": 83400 }, { "epoch": 1.6667998880940011, "grad_norm": 0.11430591344833374, "learning_rate": 5.061073522603724e-05, "loss": 0.0297, "step": 83410 }, { "epoch": 1.6669997202350026, "grad_norm": 0.283086359500885, "learning_rate": 5.0600401797659505e-05, "loss": 0.0284, "step": 83420 }, { "epoch": 1.6671995523760041, "grad_norm": 0.0951775386929512, "learning_rate": 5.05900683436337e-05, "loss": 0.0332, "step": 83430 }, { "epoch": 1.6673993845170059, "grad_norm": 0.12286680936813354, "learning_rate": 5.057973486440126e-05, "loss": 0.0337, "step": 83440 }, { "epoch": 1.6675992166580071, "grad_norm": 0.15117625892162323, "learning_rate": 5.0569401360403615e-05, "loss": 0.0385, "step": 83450 }, { "epoch": 1.6677990487990089, "grad_norm": 0.12555931508541107, "learning_rate": 5.05590678320822e-05, "loss": 0.0309, "step": 83460 }, { "epoch": 1.6679988809400104, "grad_norm": 0.1422201544046402, "learning_rate": 5.054873427987841e-05, "loss": 0.0328, "step": 83470 }, { "epoch": 1.6681987130810119, "grad_norm": 0.09467712789773941, "learning_rate": 5.0538400704233727e-05, "loss": 0.0256, "step": 83480 }, { "epoch": 1.6683985452220136, "grad_norm": 0.1288188248872757, "learning_rate": 5.0528067105589526e-05, "loss": 0.0249, "step": 83490 }, { "epoch": 1.668598377363015, "grad_norm": 0.19940698146820068, "learning_rate": 5.05177334843873e-05, "loss": 0.0302, "step": 83500 }, { "epoch": 1.6687982095040166, "grad_norm": 0.15220127999782562, "learning_rate": 5.050739984106843e-05, "loss": 0.0289, "step": 83510 }, { "epoch": 1.6689980416450183, "grad_norm": 0.12770108878612518, "learning_rate": 5.049706617607437e-05, "loss": 0.0238, "step": 83520 }, { "epoch": 1.6691978737860196, "grad_norm": 0.15574154257774353, "learning_rate": 5.0486732489846566e-05, "loss": 0.0363, "step": 83530 }, { "epoch": 1.6693977059270213, "grad_norm": 0.12573106586933136, "learning_rate": 5.047639878282643e-05, "loss": 0.0257, "step": 83540 }, { "epoch": 1.669597538068023, "grad_norm": 0.1152656227350235, "learning_rate": 5.0466065055455436e-05, "loss": 0.0289, "step": 83550 }, { "epoch": 1.6697973702090243, "grad_norm": 0.12955887615680695, "learning_rate": 5.045573130817498e-05, "loss": 0.0341, "step": 83560 }, { "epoch": 1.669997202350026, "grad_norm": 0.14434684813022614, "learning_rate": 5.044539754142653e-05, "loss": 0.0309, "step": 83570 }, { "epoch": 1.6701970344910275, "grad_norm": 0.10241110622882843, "learning_rate": 5.0435063755651516e-05, "loss": 0.0276, "step": 83580 }, { "epoch": 1.670396866632029, "grad_norm": 0.1292456090450287, "learning_rate": 5.042472995129136e-05, "loss": 0.0393, "step": 83590 }, { "epoch": 1.6705966987730307, "grad_norm": 0.12183267623186111, "learning_rate": 5.0414396128787537e-05, "loss": 0.0315, "step": 83600 }, { "epoch": 1.6707965309140322, "grad_norm": 0.13035953044891357, "learning_rate": 5.040406228858148e-05, "loss": 0.0315, "step": 83610 }, { "epoch": 1.6709963630550337, "grad_norm": 0.12007465958595276, "learning_rate": 5.0393728431114605e-05, "loss": 0.0377, "step": 83620 }, { "epoch": 1.6711961951960355, "grad_norm": 0.10064604133367538, "learning_rate": 5.038339455682839e-05, "loss": 0.0289, "step": 83630 }, { "epoch": 1.6713960273370367, "grad_norm": 0.09799520671367645, "learning_rate": 5.037306066616425e-05, "loss": 0.0283, "step": 83640 }, { "epoch": 1.6715958594780385, "grad_norm": 0.15008559823036194, "learning_rate": 5.036272675956365e-05, "loss": 0.038, "step": 83650 }, { "epoch": 1.67179569161904, "grad_norm": 0.1385466456413269, "learning_rate": 5.035239283746804e-05, "loss": 0.031, "step": 83660 }, { "epoch": 1.6719955237600415, "grad_norm": 0.1454690396785736, "learning_rate": 5.034205890031883e-05, "loss": 0.0328, "step": 83670 }, { "epoch": 1.6721953559010432, "grad_norm": 0.11047964543104172, "learning_rate": 5.033172494855753e-05, "loss": 0.0239, "step": 83680 }, { "epoch": 1.6723951880420447, "grad_norm": 0.139224573969841, "learning_rate": 5.03213909826255e-05, "loss": 0.0234, "step": 83690 }, { "epoch": 1.6725950201830462, "grad_norm": 0.13084344565868378, "learning_rate": 5.031105700296428e-05, "loss": 0.0323, "step": 83700 }, { "epoch": 1.672794852324048, "grad_norm": 0.12339358776807785, "learning_rate": 5.030072301001526e-05, "loss": 0.0315, "step": 83710 }, { "epoch": 1.6729946844650494, "grad_norm": 0.12970101833343506, "learning_rate": 5.02903890042199e-05, "loss": 0.0256, "step": 83720 }, { "epoch": 1.673194516606051, "grad_norm": 0.1131715252995491, "learning_rate": 5.028005498601968e-05, "loss": 0.0364, "step": 83730 }, { "epoch": 1.6733943487470526, "grad_norm": 0.11795757710933685, "learning_rate": 5.026972095585599e-05, "loss": 0.0292, "step": 83740 }, { "epoch": 1.673594180888054, "grad_norm": 0.11311042308807373, "learning_rate": 5.025938691417035e-05, "loss": 0.0272, "step": 83750 }, { "epoch": 1.6737940130290556, "grad_norm": 0.08778367936611176, "learning_rate": 5.024905286140416e-05, "loss": 0.0293, "step": 83760 }, { "epoch": 1.6739938451700571, "grad_norm": 0.12899312376976013, "learning_rate": 5.023871879799889e-05, "loss": 0.0316, "step": 83770 }, { "epoch": 1.6741936773110586, "grad_norm": 0.13974927365779877, "learning_rate": 5.0228384724395985e-05, "loss": 0.0298, "step": 83780 }, { "epoch": 1.6743935094520603, "grad_norm": 0.12356489151716232, "learning_rate": 5.021805064103691e-05, "loss": 0.0286, "step": 83790 }, { "epoch": 1.6745933415930618, "grad_norm": 0.09998226165771484, "learning_rate": 5.0207716548363114e-05, "loss": 0.03, "step": 83800 }, { "epoch": 1.6747931737340633, "grad_norm": 0.12247934192419052, "learning_rate": 5.019738244681605e-05, "loss": 0.0246, "step": 83810 }, { "epoch": 1.674993005875065, "grad_norm": 0.146646648645401, "learning_rate": 5.018704833683716e-05, "loss": 0.032, "step": 83820 }, { "epoch": 1.6751928380160666, "grad_norm": 0.12420675158500671, "learning_rate": 5.017671421886793e-05, "loss": 0.035, "step": 83830 }, { "epoch": 1.675392670157068, "grad_norm": 0.1905272752046585, "learning_rate": 5.016638009334977e-05, "loss": 0.0381, "step": 83840 }, { "epoch": 1.6755925022980698, "grad_norm": 0.1392621546983719, "learning_rate": 5.0156045960724164e-05, "loss": 0.0364, "step": 83850 }, { "epoch": 1.675792334439071, "grad_norm": 0.11837561428546906, "learning_rate": 5.0145711821432576e-05, "loss": 0.0312, "step": 83860 }, { "epoch": 1.6759921665800728, "grad_norm": 0.12714552879333496, "learning_rate": 5.013537767591643e-05, "loss": 0.0276, "step": 83870 }, { "epoch": 1.6761919987210743, "grad_norm": 0.13470853865146637, "learning_rate": 5.012504352461721e-05, "loss": 0.0304, "step": 83880 }, { "epoch": 1.6763918308620758, "grad_norm": 0.11584179848432541, "learning_rate": 5.0114709367976355e-05, "loss": 0.0302, "step": 83890 }, { "epoch": 1.6765916630030775, "grad_norm": 0.10883449018001556, "learning_rate": 5.010437520643532e-05, "loss": 0.0365, "step": 83900 }, { "epoch": 1.676791495144079, "grad_norm": 0.13101962208747864, "learning_rate": 5.009404104043557e-05, "loss": 0.0323, "step": 83910 }, { "epoch": 1.6769913272850805, "grad_norm": 0.12291150540113449, "learning_rate": 5.008370687041857e-05, "loss": 0.03, "step": 83920 }, { "epoch": 1.6771911594260822, "grad_norm": 0.13728418946266174, "learning_rate": 5.007337269682576e-05, "loss": 0.0345, "step": 83930 }, { "epoch": 1.6773909915670835, "grad_norm": 0.11590272933244705, "learning_rate": 5.0063038520098595e-05, "loss": 0.0243, "step": 83940 }, { "epoch": 1.6775908237080852, "grad_norm": 0.1280115842819214, "learning_rate": 5.0052704340678556e-05, "loss": 0.0299, "step": 83950 }, { "epoch": 1.677790655849087, "grad_norm": 0.18426138162612915, "learning_rate": 5.0042370159007076e-05, "loss": 0.0296, "step": 83960 }, { "epoch": 1.6779904879900882, "grad_norm": 0.12587302923202515, "learning_rate": 5.003203597552562e-05, "loss": 0.0325, "step": 83970 }, { "epoch": 1.67819032013109, "grad_norm": 0.10890621691942215, "learning_rate": 5.0021701790675656e-05, "loss": 0.0255, "step": 83980 }, { "epoch": 1.6783901522720914, "grad_norm": 0.09505847841501236, "learning_rate": 5.0011367604898615e-05, "loss": 0.0246, "step": 83990 }, { "epoch": 1.678589984413093, "grad_norm": 0.14613452553749084, "learning_rate": 5.0001033418635986e-05, "loss": 0.0341, "step": 84000 }, { "epoch": 1.6787898165540946, "grad_norm": 0.12630127370357513, "learning_rate": 4.9990699232329204e-05, "loss": 0.0295, "step": 84010 }, { "epoch": 1.6789896486950961, "grad_norm": 0.14036032557487488, "learning_rate": 4.998036504641972e-05, "loss": 0.0378, "step": 84020 }, { "epoch": 1.6791894808360976, "grad_norm": 0.10597807168960571, "learning_rate": 4.9970030861349023e-05, "loss": 0.033, "step": 84030 }, { "epoch": 1.6793893129770994, "grad_norm": 0.1252729743719101, "learning_rate": 4.9959696677558546e-05, "loss": 0.0305, "step": 84040 }, { "epoch": 1.6795891451181006, "grad_norm": 0.11692946404218674, "learning_rate": 4.9949362495489774e-05, "loss": 0.033, "step": 84050 }, { "epoch": 1.6797889772591024, "grad_norm": 0.1711389720439911, "learning_rate": 4.993902831558411e-05, "loss": 0.0275, "step": 84060 }, { "epoch": 1.6799888094001039, "grad_norm": 0.13997292518615723, "learning_rate": 4.9928694138283054e-05, "loss": 0.0329, "step": 84070 }, { "epoch": 1.6801886415411054, "grad_norm": 0.1688953936100006, "learning_rate": 4.9918359964028066e-05, "loss": 0.036, "step": 84080 }, { "epoch": 1.680388473682107, "grad_norm": 0.12818579375743866, "learning_rate": 4.9908025793260584e-05, "loss": 0.0262, "step": 84090 }, { "epoch": 1.6805883058231086, "grad_norm": 0.13492685556411743, "learning_rate": 4.9897691626422075e-05, "loss": 0.0281, "step": 84100 }, { "epoch": 1.68078813796411, "grad_norm": 0.08584278076887131, "learning_rate": 4.988735746395398e-05, "loss": 0.0269, "step": 84110 }, { "epoch": 1.6809879701051118, "grad_norm": 0.1310300976037979, "learning_rate": 4.9877023306297766e-05, "loss": 0.0292, "step": 84120 }, { "epoch": 1.6811878022461133, "grad_norm": 0.13573884963989258, "learning_rate": 4.986668915389491e-05, "loss": 0.0408, "step": 84130 }, { "epoch": 1.6813876343871148, "grad_norm": 0.11470097303390503, "learning_rate": 4.9856355007186833e-05, "loss": 0.033, "step": 84140 }, { "epoch": 1.6815874665281165, "grad_norm": 0.13862046599388123, "learning_rate": 4.9846020866615e-05, "loss": 0.0265, "step": 84150 }, { "epoch": 1.6817872986691178, "grad_norm": 0.11014049500226974, "learning_rate": 4.983568673262089e-05, "loss": 0.0269, "step": 84160 }, { "epoch": 1.6819871308101195, "grad_norm": 0.10137839615345001, "learning_rate": 4.982535260564593e-05, "loss": 0.0293, "step": 84170 }, { "epoch": 1.682186962951121, "grad_norm": 0.12483146786689758, "learning_rate": 4.98150184861316e-05, "loss": 0.0309, "step": 84180 }, { "epoch": 1.6823867950921225, "grad_norm": 0.14633499085903168, "learning_rate": 4.980468437451933e-05, "loss": 0.0232, "step": 84190 }, { "epoch": 1.6825866272331242, "grad_norm": 0.12490396201610565, "learning_rate": 4.979435027125057e-05, "loss": 0.0309, "step": 84200 }, { "epoch": 1.6827864593741257, "grad_norm": 0.12418580800294876, "learning_rate": 4.978401617676681e-05, "loss": 0.0313, "step": 84210 }, { "epoch": 1.6829862915151272, "grad_norm": 0.10985247790813446, "learning_rate": 4.9773682091509474e-05, "loss": 0.0219, "step": 84220 }, { "epoch": 1.683186123656129, "grad_norm": 0.13831017911434174, "learning_rate": 4.9763348015920035e-05, "loss": 0.025, "step": 84230 }, { "epoch": 1.6833859557971305, "grad_norm": 0.10743337124586105, "learning_rate": 4.9753013950439914e-05, "loss": 0.036, "step": 84240 }, { "epoch": 1.683585787938132, "grad_norm": 0.12196541577577591, "learning_rate": 4.974267989551059e-05, "loss": 0.028, "step": 84250 }, { "epoch": 1.6837856200791337, "grad_norm": 0.10439562797546387, "learning_rate": 4.973234585157352e-05, "loss": 0.021, "step": 84260 }, { "epoch": 1.683985452220135, "grad_norm": 0.09552045166492462, "learning_rate": 4.972201181907013e-05, "loss": 0.0241, "step": 84270 }, { "epoch": 1.6841852843611367, "grad_norm": 0.10714024305343628, "learning_rate": 4.9711677798441904e-05, "loss": 0.0278, "step": 84280 }, { "epoch": 1.6843851165021382, "grad_norm": 0.11337187886238098, "learning_rate": 4.9701343790130225e-05, "loss": 0.0294, "step": 84290 }, { "epoch": 1.6845849486431397, "grad_norm": 0.164051815867424, "learning_rate": 4.969100979457664e-05, "loss": 0.0356, "step": 84300 }, { "epoch": 1.6847847807841414, "grad_norm": 0.11470622569322586, "learning_rate": 4.968067581222253e-05, "loss": 0.0304, "step": 84310 }, { "epoch": 1.684984612925143, "grad_norm": 0.14179283380508423, "learning_rate": 4.9670341843509375e-05, "loss": 0.0225, "step": 84320 }, { "epoch": 1.6851844450661444, "grad_norm": 0.11296184360980988, "learning_rate": 4.9660007888878585e-05, "loss": 0.0282, "step": 84330 }, { "epoch": 1.6853842772071461, "grad_norm": 0.13620847463607788, "learning_rate": 4.9649673948771646e-05, "loss": 0.0318, "step": 84340 }, { "epoch": 1.6855841093481474, "grad_norm": 0.12522803246974945, "learning_rate": 4.9639340023630004e-05, "loss": 0.0268, "step": 84350 }, { "epoch": 1.6857839414891491, "grad_norm": 0.13278333842754364, "learning_rate": 4.962900611389509e-05, "loss": 0.0296, "step": 84360 }, { "epoch": 1.6859837736301508, "grad_norm": 0.12346848100423813, "learning_rate": 4.9618672220008355e-05, "loss": 0.0239, "step": 84370 }, { "epoch": 1.6861836057711521, "grad_norm": 0.1332838237285614, "learning_rate": 4.960833834241122e-05, "loss": 0.0317, "step": 84380 }, { "epoch": 1.6863834379121538, "grad_norm": 0.15489374101161957, "learning_rate": 4.959800448154517e-05, "loss": 0.0257, "step": 84390 }, { "epoch": 1.6865832700531553, "grad_norm": 0.11204178631305695, "learning_rate": 4.958767063785163e-05, "loss": 0.0342, "step": 84400 }, { "epoch": 1.6867831021941568, "grad_norm": 0.09926389902830124, "learning_rate": 4.957733681177204e-05, "loss": 0.0231, "step": 84410 }, { "epoch": 1.6869829343351586, "grad_norm": 0.12083607167005539, "learning_rate": 4.956700300374784e-05, "loss": 0.0294, "step": 84420 }, { "epoch": 1.68718276647616, "grad_norm": 0.10671902447938919, "learning_rate": 4.955666921422049e-05, "loss": 0.029, "step": 84430 }, { "epoch": 1.6873825986171616, "grad_norm": 0.11869698017835617, "learning_rate": 4.9546335443631405e-05, "loss": 0.0219, "step": 84440 }, { "epoch": 1.6875824307581633, "grad_norm": 0.11268573999404907, "learning_rate": 4.953600169242205e-05, "loss": 0.0217, "step": 84450 }, { "epoch": 1.6877822628991646, "grad_norm": 0.14178712666034698, "learning_rate": 4.952566796103383e-05, "loss": 0.0255, "step": 84460 }, { "epoch": 1.6879820950401663, "grad_norm": 0.10509690642356873, "learning_rate": 4.9515334249908204e-05, "loss": 0.0362, "step": 84470 }, { "epoch": 1.6881819271811678, "grad_norm": 0.11967767030000687, "learning_rate": 4.950500055948663e-05, "loss": 0.0301, "step": 84480 }, { "epoch": 1.6883817593221693, "grad_norm": 0.1270846128463745, "learning_rate": 4.9494666890210505e-05, "loss": 0.0353, "step": 84490 }, { "epoch": 1.688581591463171, "grad_norm": 0.15264903008937836, "learning_rate": 4.9484333242521295e-05, "loss": 0.0308, "step": 84500 }, { "epoch": 1.6887814236041725, "grad_norm": 0.155757874250412, "learning_rate": 4.947399961686041e-05, "loss": 0.0283, "step": 84510 }, { "epoch": 1.688981255745174, "grad_norm": 0.1412920206785202, "learning_rate": 4.946366601366929e-05, "loss": 0.0309, "step": 84520 }, { "epoch": 1.6891810878861757, "grad_norm": 0.14595502614974976, "learning_rate": 4.945333243338939e-05, "loss": 0.0274, "step": 84530 }, { "epoch": 1.6893809200271772, "grad_norm": 0.09220565110445023, "learning_rate": 4.944299887646211e-05, "loss": 0.0237, "step": 84540 }, { "epoch": 1.6895807521681787, "grad_norm": 0.1495269536972046, "learning_rate": 4.943266534332892e-05, "loss": 0.0334, "step": 84550 }, { "epoch": 1.6897805843091804, "grad_norm": 0.10476212203502655, "learning_rate": 4.942233183443119e-05, "loss": 0.0349, "step": 84560 }, { "epoch": 1.6899804164501817, "grad_norm": 0.16043496131896973, "learning_rate": 4.9411998350210395e-05, "loss": 0.0339, "step": 84570 }, { "epoch": 1.6901802485911834, "grad_norm": 0.13862508535385132, "learning_rate": 4.940166489110796e-05, "loss": 0.0265, "step": 84580 }, { "epoch": 1.690380080732185, "grad_norm": 0.1068514883518219, "learning_rate": 4.9391331457565295e-05, "loss": 0.0304, "step": 84590 }, { "epoch": 1.6905799128731864, "grad_norm": 0.15511898696422577, "learning_rate": 4.9380998050023816e-05, "loss": 0.0297, "step": 84600 }, { "epoch": 1.6907797450141882, "grad_norm": 0.14741270244121552, "learning_rate": 4.937066466892499e-05, "loss": 0.0237, "step": 84610 }, { "epoch": 1.6909795771551897, "grad_norm": 0.10864201188087463, "learning_rate": 4.93603313147102e-05, "loss": 0.0275, "step": 84620 }, { "epoch": 1.6911794092961911, "grad_norm": 0.09983135759830475, "learning_rate": 4.9349997987820894e-05, "loss": 0.0223, "step": 84630 }, { "epoch": 1.6913792414371929, "grad_norm": 0.16392628848552704, "learning_rate": 4.933966468869847e-05, "loss": 0.0339, "step": 84640 }, { "epoch": 1.6915790735781944, "grad_norm": 0.1556890457868576, "learning_rate": 4.9329331417784344e-05, "loss": 0.0317, "step": 84650 }, { "epoch": 1.6917789057191959, "grad_norm": 0.14689281582832336, "learning_rate": 4.9318998175519966e-05, "loss": 0.0319, "step": 84660 }, { "epoch": 1.6919787378601976, "grad_norm": 0.10921797156333923, "learning_rate": 4.930866496234673e-05, "loss": 0.0293, "step": 84670 }, { "epoch": 1.6921785700011989, "grad_norm": 0.11049497127532959, "learning_rate": 4.9298331778706066e-05, "loss": 0.0327, "step": 84680 }, { "epoch": 1.6923784021422006, "grad_norm": 0.15019318461418152, "learning_rate": 4.928799862503934e-05, "loss": 0.0239, "step": 84690 }, { "epoch": 1.692578234283202, "grad_norm": 0.12033481895923615, "learning_rate": 4.9277665501788046e-05, "loss": 0.0274, "step": 84700 }, { "epoch": 1.6927780664242036, "grad_norm": 0.09639584273099899, "learning_rate": 4.9267332409393544e-05, "loss": 0.0281, "step": 84710 }, { "epoch": 1.6929778985652053, "grad_norm": 0.10349198430776596, "learning_rate": 4.925699934829726e-05, "loss": 0.0387, "step": 84720 }, { "epoch": 1.6931777307062068, "grad_norm": 0.11999207735061646, "learning_rate": 4.924666631894059e-05, "loss": 0.0249, "step": 84730 }, { "epoch": 1.6933775628472083, "grad_norm": 0.12465182691812515, "learning_rate": 4.9236333321764945e-05, "loss": 0.03, "step": 84740 }, { "epoch": 1.69357739498821, "grad_norm": 0.13907988369464874, "learning_rate": 4.922600035721176e-05, "loss": 0.0356, "step": 84750 }, { "epoch": 1.6937772271292113, "grad_norm": 0.1142820417881012, "learning_rate": 4.92156674257224e-05, "loss": 0.04, "step": 84760 }, { "epoch": 1.693977059270213, "grad_norm": 0.15192832052707672, "learning_rate": 4.92053345277383e-05, "loss": 0.036, "step": 84770 }, { "epoch": 1.6941768914112145, "grad_norm": 0.14141571521759033, "learning_rate": 4.9195001663700826e-05, "loss": 0.0227, "step": 84780 }, { "epoch": 1.694376723552216, "grad_norm": 0.16386093199253082, "learning_rate": 4.918466883405142e-05, "loss": 0.0271, "step": 84790 }, { "epoch": 1.6945765556932177, "grad_norm": 0.13461925089359283, "learning_rate": 4.917433603923147e-05, "loss": 0.0249, "step": 84800 }, { "epoch": 1.6947763878342192, "grad_norm": 0.16083478927612305, "learning_rate": 4.916400327968235e-05, "loss": 0.0264, "step": 84810 }, { "epoch": 1.6949762199752207, "grad_norm": 0.11569489538669586, "learning_rate": 4.915367055584549e-05, "loss": 0.0285, "step": 84820 }, { "epoch": 1.6951760521162225, "grad_norm": 0.14878983795642853, "learning_rate": 4.9143337868162254e-05, "loss": 0.03, "step": 84830 }, { "epoch": 1.695375884257224, "grad_norm": 0.12644487619400024, "learning_rate": 4.913300521707406e-05, "loss": 0.0322, "step": 84840 }, { "epoch": 1.6955757163982255, "grad_norm": 0.22474709153175354, "learning_rate": 4.912267260302229e-05, "loss": 0.0246, "step": 84850 }, { "epoch": 1.6957755485392272, "grad_norm": 0.13468128442764282, "learning_rate": 4.9112340026448336e-05, "loss": 0.0244, "step": 84860 }, { "epoch": 1.6959753806802285, "grad_norm": 0.12448015064001083, "learning_rate": 4.9102007487793566e-05, "loss": 0.0325, "step": 84870 }, { "epoch": 1.6961752128212302, "grad_norm": 0.11851964890956879, "learning_rate": 4.9091674987499414e-05, "loss": 0.0315, "step": 84880 }, { "epoch": 1.6963750449622317, "grad_norm": 0.12992344796657562, "learning_rate": 4.9081342526007225e-05, "loss": 0.0277, "step": 84890 }, { "epoch": 1.6965748771032332, "grad_norm": 0.12492641061544418, "learning_rate": 4.9071010103758404e-05, "loss": 0.0291, "step": 84900 }, { "epoch": 1.696774709244235, "grad_norm": 0.14701604843139648, "learning_rate": 4.906067772119432e-05, "loss": 0.0297, "step": 84910 }, { "epoch": 1.6969745413852364, "grad_norm": 0.14447669684886932, "learning_rate": 4.905034537875634e-05, "loss": 0.0337, "step": 84920 }, { "epoch": 1.697174373526238, "grad_norm": 0.16365070641040802, "learning_rate": 4.9040013076885886e-05, "loss": 0.0275, "step": 84930 }, { "epoch": 1.6973742056672396, "grad_norm": 0.12065194547176361, "learning_rate": 4.90296808160243e-05, "loss": 0.032, "step": 84940 }, { "epoch": 1.6975740378082411, "grad_norm": 0.09743781387805939, "learning_rate": 4.901934859661298e-05, "loss": 0.026, "step": 84950 }, { "epoch": 1.6977738699492426, "grad_norm": 0.12144692987203598, "learning_rate": 4.9009016419093256e-05, "loss": 0.037, "step": 84960 }, { "epoch": 1.6979737020902443, "grad_norm": 0.12863671779632568, "learning_rate": 4.899868428390655e-05, "loss": 0.0211, "step": 84970 }, { "epoch": 1.6981735342312456, "grad_norm": 0.09279360622167587, "learning_rate": 4.8988352191494216e-05, "loss": 0.0339, "step": 84980 }, { "epoch": 1.6983733663722473, "grad_norm": 0.181560218334198, "learning_rate": 4.897802014229762e-05, "loss": 0.0228, "step": 84990 }, { "epoch": 1.6985731985132488, "grad_norm": 0.1528172791004181, "learning_rate": 4.896768813675812e-05, "loss": 0.0281, "step": 85000 }, { "epoch": 1.6987730306542503, "grad_norm": 0.16090509295463562, "learning_rate": 4.895735617531707e-05, "loss": 0.0238, "step": 85010 }, { "epoch": 1.698972862795252, "grad_norm": 0.14037129282951355, "learning_rate": 4.8947024258415864e-05, "loss": 0.0299, "step": 85020 }, { "epoch": 1.6991726949362536, "grad_norm": 0.10288181900978088, "learning_rate": 4.8936692386495855e-05, "loss": 0.0227, "step": 85030 }, { "epoch": 1.699372527077255, "grad_norm": 0.11120719462633133, "learning_rate": 4.892636055999839e-05, "loss": 0.0242, "step": 85040 }, { "epoch": 1.6995723592182568, "grad_norm": 0.15756092965602875, "learning_rate": 4.891602877936481e-05, "loss": 0.0331, "step": 85050 }, { "epoch": 1.6997721913592583, "grad_norm": 0.13637323677539825, "learning_rate": 4.890569704503651e-05, "loss": 0.0332, "step": 85060 }, { "epoch": 1.6999720235002598, "grad_norm": 0.11537330597639084, "learning_rate": 4.889536535745481e-05, "loss": 0.021, "step": 85070 }, { "epoch": 1.7001718556412615, "grad_norm": 0.12527155876159668, "learning_rate": 4.888503371706109e-05, "loss": 0.0256, "step": 85080 }, { "epoch": 1.7003716877822628, "grad_norm": 0.14871527254581451, "learning_rate": 4.887470212429665e-05, "loss": 0.0297, "step": 85090 }, { "epoch": 1.7005715199232645, "grad_norm": 0.12980686128139496, "learning_rate": 4.88643705796029e-05, "loss": 0.025, "step": 85100 }, { "epoch": 1.700771352064266, "grad_norm": 0.13499139249324799, "learning_rate": 4.885403908342114e-05, "loss": 0.0286, "step": 85110 }, { "epoch": 1.7009711842052675, "grad_norm": 0.08553117513656616, "learning_rate": 4.884370763619273e-05, "loss": 0.0283, "step": 85120 }, { "epoch": 1.7011710163462692, "grad_norm": 0.12744493782520294, "learning_rate": 4.8833376238359e-05, "loss": 0.0243, "step": 85130 }, { "epoch": 1.7013708484872707, "grad_norm": 0.13268905878067017, "learning_rate": 4.882304489036129e-05, "loss": 0.0309, "step": 85140 }, { "epoch": 1.7015706806282722, "grad_norm": 0.11925002187490463, "learning_rate": 4.881271359264095e-05, "loss": 0.0278, "step": 85150 }, { "epoch": 1.701770512769274, "grad_norm": 0.12415458261966705, "learning_rate": 4.88023823456393e-05, "loss": 0.0287, "step": 85160 }, { "epoch": 1.7019703449102752, "grad_norm": 0.12789325416088104, "learning_rate": 4.8792051149797684e-05, "loss": 0.0377, "step": 85170 }, { "epoch": 1.702170177051277, "grad_norm": 0.14776049554347992, "learning_rate": 4.87817200055574e-05, "loss": 0.0242, "step": 85180 }, { "epoch": 1.7023700091922784, "grad_norm": 0.13241256773471832, "learning_rate": 4.8771388913359814e-05, "loss": 0.0246, "step": 85190 }, { "epoch": 1.70256984133328, "grad_norm": 0.14082428812980652, "learning_rate": 4.876105787364625e-05, "loss": 0.0284, "step": 85200 }, { "epoch": 1.7027696734742817, "grad_norm": 0.13174965977668762, "learning_rate": 4.8750726886858e-05, "loss": 0.0368, "step": 85210 }, { "epoch": 1.7029695056152832, "grad_norm": 0.14043134450912476, "learning_rate": 4.874039595343642e-05, "loss": 0.0294, "step": 85220 }, { "epoch": 1.7031693377562847, "grad_norm": 0.1396304965019226, "learning_rate": 4.8730065073822784e-05, "loss": 0.0317, "step": 85230 }, { "epoch": 1.7033691698972864, "grad_norm": 0.11396114528179169, "learning_rate": 4.8719734248458447e-05, "loss": 0.033, "step": 85240 }, { "epoch": 1.7035690020382879, "grad_norm": 0.1175268217921257, "learning_rate": 4.870940347778473e-05, "loss": 0.0266, "step": 85250 }, { "epoch": 1.7037688341792894, "grad_norm": 0.12103051692247391, "learning_rate": 4.86990727622429e-05, "loss": 0.0248, "step": 85260 }, { "epoch": 1.703968666320291, "grad_norm": 0.11324993520975113, "learning_rate": 4.868874210227429e-05, "loss": 0.023, "step": 85270 }, { "epoch": 1.7041684984612924, "grad_norm": 0.1578129678964615, "learning_rate": 4.867841149832023e-05, "loss": 0.0285, "step": 85280 }, { "epoch": 1.704368330602294, "grad_norm": 0.1152220368385315, "learning_rate": 4.866808095082199e-05, "loss": 0.0291, "step": 85290 }, { "epoch": 1.7045681627432956, "grad_norm": 0.15799963474273682, "learning_rate": 4.8657750460220895e-05, "loss": 0.0316, "step": 85300 }, { "epoch": 1.704767994884297, "grad_norm": 0.13194319605827332, "learning_rate": 4.864742002695823e-05, "loss": 0.0291, "step": 85310 }, { "epoch": 1.7049678270252988, "grad_norm": 0.1418496072292328, "learning_rate": 4.8637089651475285e-05, "loss": 0.0265, "step": 85320 }, { "epoch": 1.7051676591663003, "grad_norm": 0.09190592914819717, "learning_rate": 4.862675933421338e-05, "loss": 0.0282, "step": 85330 }, { "epoch": 1.7053674913073018, "grad_norm": 0.11436347663402557, "learning_rate": 4.861642907561378e-05, "loss": 0.0347, "step": 85340 }, { "epoch": 1.7055673234483035, "grad_norm": 0.1459687054157257, "learning_rate": 4.860609887611781e-05, "loss": 0.0332, "step": 85350 }, { "epoch": 1.705767155589305, "grad_norm": 0.13584193587303162, "learning_rate": 4.8595768736166715e-05, "loss": 0.0288, "step": 85360 }, { "epoch": 1.7059669877303065, "grad_norm": 0.11772971600294113, "learning_rate": 4.8585438656201806e-05, "loss": 0.0297, "step": 85370 }, { "epoch": 1.7061668198713082, "grad_norm": 0.13044331967830658, "learning_rate": 4.857510863666436e-05, "loss": 0.0332, "step": 85380 }, { "epoch": 1.7063666520123095, "grad_norm": 0.11600900441408157, "learning_rate": 4.856477867799566e-05, "loss": 0.03, "step": 85390 }, { "epoch": 1.7065664841533112, "grad_norm": 0.1353563517332077, "learning_rate": 4.855444878063698e-05, "loss": 0.0339, "step": 85400 }, { "epoch": 1.7067663162943127, "grad_norm": 0.12555696070194244, "learning_rate": 4.8544118945029575e-05, "loss": 0.0277, "step": 85410 }, { "epoch": 1.7069661484353142, "grad_norm": 0.14096699655056, "learning_rate": 4.853378917161474e-05, "loss": 0.0275, "step": 85420 }, { "epoch": 1.707165980576316, "grad_norm": 0.1420421451330185, "learning_rate": 4.8523459460833755e-05, "loss": 0.0294, "step": 85430 }, { "epoch": 1.7073658127173175, "grad_norm": 0.1600247174501419, "learning_rate": 4.851312981312785e-05, "loss": 0.0423, "step": 85440 }, { "epoch": 1.707565644858319, "grad_norm": 0.12199582904577255, "learning_rate": 4.8502800228938314e-05, "loss": 0.0244, "step": 85450 }, { "epoch": 1.7077654769993207, "grad_norm": 0.22997625172138214, "learning_rate": 4.84924707087064e-05, "loss": 0.0232, "step": 85460 }, { "epoch": 1.707965309140322, "grad_norm": 0.1316518485546112, "learning_rate": 4.848214125287338e-05, "loss": 0.0317, "step": 85470 }, { "epoch": 1.7081651412813237, "grad_norm": 0.14601773023605347, "learning_rate": 4.847181186188048e-05, "loss": 0.0294, "step": 85480 }, { "epoch": 1.7083649734223254, "grad_norm": 0.16020925343036652, "learning_rate": 4.8461482536169e-05, "loss": 0.0309, "step": 85490 }, { "epoch": 1.7085648055633267, "grad_norm": 0.13237452507019043, "learning_rate": 4.845115327618012e-05, "loss": 0.0308, "step": 85500 }, { "epoch": 1.7087646377043284, "grad_norm": 0.12208041548728943, "learning_rate": 4.844082408235514e-05, "loss": 0.0258, "step": 85510 }, { "epoch": 1.70896446984533, "grad_norm": 0.10529865324497223, "learning_rate": 4.843049495513531e-05, "loss": 0.0244, "step": 85520 }, { "epoch": 1.7091643019863314, "grad_norm": 0.12221917510032654, "learning_rate": 4.842016589496183e-05, "loss": 0.027, "step": 85530 }, { "epoch": 1.7093641341273331, "grad_norm": 0.11858358234167099, "learning_rate": 4.8409836902275966e-05, "loss": 0.0263, "step": 85540 }, { "epoch": 1.7095639662683346, "grad_norm": 0.105145663022995, "learning_rate": 4.839950797751895e-05, "loss": 0.0267, "step": 85550 }, { "epoch": 1.7097637984093361, "grad_norm": 0.14850550889968872, "learning_rate": 4.838917912113202e-05, "loss": 0.0261, "step": 85560 }, { "epoch": 1.7099636305503378, "grad_norm": 0.12171906977891922, "learning_rate": 4.83788503335564e-05, "loss": 0.0263, "step": 85570 }, { "epoch": 1.7101634626913391, "grad_norm": 0.19242887198925018, "learning_rate": 4.836852161523331e-05, "loss": 0.0354, "step": 85580 }, { "epoch": 1.7103632948323408, "grad_norm": 0.15683110058307648, "learning_rate": 4.835819296660396e-05, "loss": 0.0244, "step": 85590 }, { "epoch": 1.7105631269733423, "grad_norm": 0.1284472644329071, "learning_rate": 4.834786438810962e-05, "loss": 0.0322, "step": 85600 }, { "epoch": 1.7107629591143438, "grad_norm": 0.11144483089447021, "learning_rate": 4.8337535880191467e-05, "loss": 0.0274, "step": 85610 }, { "epoch": 1.7109627912553456, "grad_norm": 0.1156918928027153, "learning_rate": 4.8327207443290735e-05, "loss": 0.0292, "step": 85620 }, { "epoch": 1.711162623396347, "grad_norm": 0.13635210692882538, "learning_rate": 4.8316879077848603e-05, "loss": 0.0307, "step": 85630 }, { "epoch": 1.7113624555373486, "grad_norm": 0.13421666622161865, "learning_rate": 4.830655078430632e-05, "loss": 0.0341, "step": 85640 }, { "epoch": 1.7115622876783503, "grad_norm": 0.11950486898422241, "learning_rate": 4.8296222563105085e-05, "loss": 0.0285, "step": 85650 }, { "epoch": 1.7117621198193518, "grad_norm": 0.1298496425151825, "learning_rate": 4.828589441468609e-05, "loss": 0.0297, "step": 85660 }, { "epoch": 1.7119619519603533, "grad_norm": 0.1682053804397583, "learning_rate": 4.8275566339490544e-05, "loss": 0.0351, "step": 85670 }, { "epoch": 1.712161784101355, "grad_norm": 0.13439339399337769, "learning_rate": 4.8265238337959604e-05, "loss": 0.0259, "step": 85680 }, { "epoch": 1.7123616162423563, "grad_norm": 0.13979403674602509, "learning_rate": 4.825491041053451e-05, "loss": 0.0243, "step": 85690 }, { "epoch": 1.712561448383358, "grad_norm": 0.10526680946350098, "learning_rate": 4.824458255765646e-05, "loss": 0.0303, "step": 85700 }, { "epoch": 1.7127612805243595, "grad_norm": 0.11538784950971603, "learning_rate": 4.82342547797666e-05, "loss": 0.0299, "step": 85710 }, { "epoch": 1.712961112665361, "grad_norm": 0.11686722934246063, "learning_rate": 4.822392707730612e-05, "loss": 0.0296, "step": 85720 }, { "epoch": 1.7131609448063627, "grad_norm": 0.12511667609214783, "learning_rate": 4.821359945071623e-05, "loss": 0.0263, "step": 85730 }, { "epoch": 1.7133607769473642, "grad_norm": 0.12101058661937714, "learning_rate": 4.8203271900438084e-05, "loss": 0.0295, "step": 85740 }, { "epoch": 1.7135606090883657, "grad_norm": 0.13844135403633118, "learning_rate": 4.819294442691288e-05, "loss": 0.0333, "step": 85750 }, { "epoch": 1.7137604412293674, "grad_norm": 0.10575947910547256, "learning_rate": 4.8182617030581754e-05, "loss": 0.025, "step": 85760 }, { "epoch": 1.713960273370369, "grad_norm": 0.10766143351793289, "learning_rate": 4.817228971188587e-05, "loss": 0.0282, "step": 85770 }, { "epoch": 1.7141601055113704, "grad_norm": 0.12745295464992523, "learning_rate": 4.8161962471266445e-05, "loss": 0.0242, "step": 85780 }, { "epoch": 1.7143599376523722, "grad_norm": 0.11052392423152924, "learning_rate": 4.815163530916459e-05, "loss": 0.0294, "step": 85790 }, { "epoch": 1.7145597697933734, "grad_norm": 0.1188197210431099, "learning_rate": 4.814130822602149e-05, "loss": 0.0313, "step": 85800 }, { "epoch": 1.7147596019343752, "grad_norm": 0.18152950704097748, "learning_rate": 4.813098122227826e-05, "loss": 0.0192, "step": 85810 }, { "epoch": 1.7149594340753767, "grad_norm": 0.10160128772258759, "learning_rate": 4.8120654298376106e-05, "loss": 0.029, "step": 85820 }, { "epoch": 1.7151592662163782, "grad_norm": 0.14300964772701263, "learning_rate": 4.8110327454756146e-05, "loss": 0.0309, "step": 85830 }, { "epoch": 1.7153590983573799, "grad_norm": 0.1814095824956894, "learning_rate": 4.8100000691859524e-05, "loss": 0.0403, "step": 85840 }, { "epoch": 1.7155589304983814, "grad_norm": 0.11783444881439209, "learning_rate": 4.808967401012738e-05, "loss": 0.0254, "step": 85850 }, { "epoch": 1.7157587626393829, "grad_norm": 0.1316400021314621, "learning_rate": 4.807934741000083e-05, "loss": 0.022, "step": 85860 }, { "epoch": 1.7159585947803846, "grad_norm": 0.1558174341917038, "learning_rate": 4.806902089192106e-05, "loss": 0.0483, "step": 85870 }, { "epoch": 1.7161584269213859, "grad_norm": 0.11802759021520615, "learning_rate": 4.805869445632916e-05, "loss": 0.0297, "step": 85880 }, { "epoch": 1.7163582590623876, "grad_norm": 0.08142442256212234, "learning_rate": 4.804836810366628e-05, "loss": 0.0287, "step": 85890 }, { "epoch": 1.7165580912033893, "grad_norm": 0.10626278817653656, "learning_rate": 4.8038041834373507e-05, "loss": 0.0224, "step": 85900 }, { "epoch": 1.7167579233443906, "grad_norm": 0.08458945155143738, "learning_rate": 4.8027715648891994e-05, "loss": 0.031, "step": 85910 }, { "epoch": 1.7169577554853923, "grad_norm": 0.12693865597248077, "learning_rate": 4.801738954766285e-05, "loss": 0.0277, "step": 85920 }, { "epoch": 1.7171575876263938, "grad_norm": 0.09724318981170654, "learning_rate": 4.800706353112717e-05, "loss": 0.037, "step": 85930 }, { "epoch": 1.7173574197673953, "grad_norm": 0.1258905827999115, "learning_rate": 4.7996737599726095e-05, "loss": 0.0284, "step": 85940 }, { "epoch": 1.717557251908397, "grad_norm": 0.11009806394577026, "learning_rate": 4.798641175390069e-05, "loss": 0.0294, "step": 85950 }, { "epoch": 1.7177570840493985, "grad_norm": 0.13337481021881104, "learning_rate": 4.797608599409207e-05, "loss": 0.0261, "step": 85960 }, { "epoch": 1.7179569161904, "grad_norm": 0.12085992842912674, "learning_rate": 4.796576032074136e-05, "loss": 0.0254, "step": 85970 }, { "epoch": 1.7181567483314018, "grad_norm": 0.16009709239006042, "learning_rate": 4.795543473428962e-05, "loss": 0.0254, "step": 85980 }, { "epoch": 1.718356580472403, "grad_norm": 0.1509544849395752, "learning_rate": 4.7945109235177946e-05, "loss": 0.0313, "step": 85990 }, { "epoch": 1.7185564126134047, "grad_norm": 0.11714958399534225, "learning_rate": 4.7934783823847454e-05, "loss": 0.0248, "step": 86000 }, { "epoch": 1.7187562447544062, "grad_norm": 0.13344615697860718, "learning_rate": 4.7924458500739183e-05, "loss": 0.0215, "step": 86010 }, { "epoch": 1.7189560768954077, "grad_norm": 0.2223621904850006, "learning_rate": 4.791413326629425e-05, "loss": 0.0384, "step": 86020 }, { "epoch": 1.7191559090364095, "grad_norm": 0.15259824693202972, "learning_rate": 4.790380812095369e-05, "loss": 0.0255, "step": 86030 }, { "epoch": 1.719355741177411, "grad_norm": 0.09482018649578094, "learning_rate": 4.78934830651586e-05, "loss": 0.0354, "step": 86040 }, { "epoch": 1.7195555733184125, "grad_norm": 0.13471448421478271, "learning_rate": 4.7883158099350054e-05, "loss": 0.0338, "step": 86050 }, { "epoch": 1.7197554054594142, "grad_norm": 0.1300434023141861, "learning_rate": 4.78728332239691e-05, "loss": 0.0251, "step": 86060 }, { "epoch": 1.7199552376004157, "grad_norm": 0.12936249375343323, "learning_rate": 4.7862508439456814e-05, "loss": 0.0327, "step": 86070 }, { "epoch": 1.7201550697414172, "grad_norm": 0.12528760731220245, "learning_rate": 4.7852183746254217e-05, "loss": 0.0394, "step": 86080 }, { "epoch": 1.720354901882419, "grad_norm": 0.1472185105085373, "learning_rate": 4.784185914480239e-05, "loss": 0.0237, "step": 86090 }, { "epoch": 1.7205547340234202, "grad_norm": 0.11268345266580582, "learning_rate": 4.78315346355424e-05, "loss": 0.0276, "step": 86100 }, { "epoch": 1.720754566164422, "grad_norm": 0.10320843011140823, "learning_rate": 4.782121021891525e-05, "loss": 0.0282, "step": 86110 }, { "epoch": 1.7209543983054234, "grad_norm": 0.09617361426353455, "learning_rate": 4.7810885895362e-05, "loss": 0.0319, "step": 86120 }, { "epoch": 1.721154230446425, "grad_norm": 0.119856096804142, "learning_rate": 4.780056166532367e-05, "loss": 0.0289, "step": 86130 }, { "epoch": 1.7213540625874266, "grad_norm": 0.11128608882427216, "learning_rate": 4.779023752924131e-05, "loss": 0.0253, "step": 86140 }, { "epoch": 1.7215538947284281, "grad_norm": 0.1502680778503418, "learning_rate": 4.777991348755596e-05, "loss": 0.0286, "step": 86150 }, { "epoch": 1.7217537268694296, "grad_norm": 0.10540342330932617, "learning_rate": 4.776958954070861e-05, "loss": 0.0241, "step": 86160 }, { "epoch": 1.7219535590104313, "grad_norm": 0.11238507926464081, "learning_rate": 4.77592656891403e-05, "loss": 0.0353, "step": 86170 }, { "epoch": 1.7221533911514328, "grad_norm": 0.14067891240119934, "learning_rate": 4.774894193329205e-05, "loss": 0.0298, "step": 86180 }, { "epoch": 1.7223532232924343, "grad_norm": 0.11799867451190948, "learning_rate": 4.773861827360486e-05, "loss": 0.0278, "step": 86190 }, { "epoch": 1.722553055433436, "grad_norm": 0.13134127855300903, "learning_rate": 4.7728294710519756e-05, "loss": 0.0245, "step": 86200 }, { "epoch": 1.7227528875744373, "grad_norm": 0.1228252425789833, "learning_rate": 4.771797124447772e-05, "loss": 0.0324, "step": 86210 }, { "epoch": 1.722952719715439, "grad_norm": 0.12819813191890717, "learning_rate": 4.7707647875919745e-05, "loss": 0.024, "step": 86220 }, { "epoch": 1.7231525518564406, "grad_norm": 0.12124611437320709, "learning_rate": 4.769732460528686e-05, "loss": 0.0331, "step": 86230 }, { "epoch": 1.723352383997442, "grad_norm": 0.11091301590204239, "learning_rate": 4.768700143302003e-05, "loss": 0.0272, "step": 86240 }, { "epoch": 1.7235522161384438, "grad_norm": 0.14889493584632874, "learning_rate": 4.767667835956026e-05, "loss": 0.0299, "step": 86250 }, { "epoch": 1.7237520482794453, "grad_norm": 0.11702115833759308, "learning_rate": 4.7666355385348497e-05, "loss": 0.0315, "step": 86260 }, { "epoch": 1.7239518804204468, "grad_norm": 0.11967498809099197, "learning_rate": 4.7656032510825763e-05, "loss": 0.0274, "step": 86270 }, { "epoch": 1.7241517125614485, "grad_norm": 0.1496405154466629, "learning_rate": 4.764570973643301e-05, "loss": 0.0238, "step": 86280 }, { "epoch": 1.7243515447024498, "grad_norm": 0.12197766453027725, "learning_rate": 4.763538706261123e-05, "loss": 0.0356, "step": 86290 }, { "epoch": 1.7245513768434515, "grad_norm": 0.14481738209724426, "learning_rate": 4.762506448980135e-05, "loss": 0.0254, "step": 86300 }, { "epoch": 1.7247512089844532, "grad_norm": 0.12950755655765533, "learning_rate": 4.761474201844435e-05, "loss": 0.0262, "step": 86310 }, { "epoch": 1.7249510411254545, "grad_norm": 0.1381780058145523, "learning_rate": 4.760441964898121e-05, "loss": 0.0318, "step": 86320 }, { "epoch": 1.7251508732664562, "grad_norm": 0.10836775600910187, "learning_rate": 4.7594097381852845e-05, "loss": 0.0293, "step": 86330 }, { "epoch": 1.7253507054074577, "grad_norm": 0.12910738587379456, "learning_rate": 4.7583775217500234e-05, "loss": 0.033, "step": 86340 }, { "epoch": 1.7255505375484592, "grad_norm": 0.11635760217905045, "learning_rate": 4.757345315636429e-05, "loss": 0.0242, "step": 86350 }, { "epoch": 1.725750369689461, "grad_norm": 0.15467755496501923, "learning_rate": 4.756313119888597e-05, "loss": 0.0276, "step": 86360 }, { "epoch": 1.7259502018304624, "grad_norm": 0.11864537000656128, "learning_rate": 4.7552809345506224e-05, "loss": 0.0251, "step": 86370 }, { "epoch": 1.726150033971464, "grad_norm": 0.14460521936416626, "learning_rate": 4.754248759666596e-05, "loss": 0.0245, "step": 86380 }, { "epoch": 1.7263498661124657, "grad_norm": 0.08462315797805786, "learning_rate": 4.7532165952806095e-05, "loss": 0.0286, "step": 86390 }, { "epoch": 1.726549698253467, "grad_norm": 0.09634813666343689, "learning_rate": 4.7521844414367586e-05, "loss": 0.0326, "step": 86400 }, { "epoch": 1.7267495303944687, "grad_norm": 0.10946785658597946, "learning_rate": 4.751152298179133e-05, "loss": 0.0278, "step": 86410 }, { "epoch": 1.7269493625354702, "grad_norm": 0.10927274823188782, "learning_rate": 4.7501201655518243e-05, "loss": 0.0218, "step": 86420 }, { "epoch": 1.7271491946764717, "grad_norm": 0.13146360218524933, "learning_rate": 4.7490880435989225e-05, "loss": 0.0329, "step": 86430 }, { "epoch": 1.7273490268174734, "grad_norm": 0.0941987857222557, "learning_rate": 4.748055932364517e-05, "loss": 0.036, "step": 86440 }, { "epoch": 1.7275488589584749, "grad_norm": 0.13832518458366394, "learning_rate": 4.747023831892702e-05, "loss": 0.0309, "step": 86450 }, { "epoch": 1.7277486910994764, "grad_norm": 0.11653710156679153, "learning_rate": 4.745991742227563e-05, "loss": 0.0293, "step": 86460 }, { "epoch": 1.727948523240478, "grad_norm": 0.11210959404706955, "learning_rate": 4.744959663413191e-05, "loss": 0.0239, "step": 86470 }, { "epoch": 1.7281483553814796, "grad_norm": 0.09659560024738312, "learning_rate": 4.7439275954936715e-05, "loss": 0.0252, "step": 86480 }, { "epoch": 1.728348187522481, "grad_norm": 0.09084486961364746, "learning_rate": 4.742895538513096e-05, "loss": 0.0265, "step": 86490 }, { "epoch": 1.7285480196634828, "grad_norm": 0.10914808511734009, "learning_rate": 4.7418634925155514e-05, "loss": 0.0351, "step": 86500 }, { "epoch": 1.728747851804484, "grad_norm": 0.10856908559799194, "learning_rate": 4.7408314575451236e-05, "loss": 0.0283, "step": 86510 }, { "epoch": 1.7289476839454858, "grad_norm": 0.1341692954301834, "learning_rate": 4.739799433645901e-05, "loss": 0.0246, "step": 86520 }, { "epoch": 1.7291475160864873, "grad_norm": 0.11278229206800461, "learning_rate": 4.738767420861966e-05, "loss": 0.0279, "step": 86530 }, { "epoch": 1.7293473482274888, "grad_norm": 0.1522076427936554, "learning_rate": 4.7377354192374076e-05, "loss": 0.0272, "step": 86540 }, { "epoch": 1.7295471803684905, "grad_norm": 0.13799554109573364, "learning_rate": 4.736703428816312e-05, "loss": 0.0303, "step": 86550 }, { "epoch": 1.729747012509492, "grad_norm": 0.10826326906681061, "learning_rate": 4.735671449642761e-05, "loss": 0.0276, "step": 86560 }, { "epoch": 1.7299468446504935, "grad_norm": 0.14043398201465607, "learning_rate": 4.734639481760839e-05, "loss": 0.0291, "step": 86570 }, { "epoch": 1.7301466767914953, "grad_norm": 0.11579757928848267, "learning_rate": 4.733607525214633e-05, "loss": 0.0296, "step": 86580 }, { "epoch": 1.7303465089324968, "grad_norm": 0.12808170914649963, "learning_rate": 4.732575580048223e-05, "loss": 0.0316, "step": 86590 }, { "epoch": 1.7305463410734983, "grad_norm": 0.10798004269599915, "learning_rate": 4.731543646305694e-05, "loss": 0.0264, "step": 86600 }, { "epoch": 1.7307461732145, "grad_norm": 0.17672088742256165, "learning_rate": 4.7305117240311265e-05, "loss": 0.0271, "step": 86610 }, { "epoch": 1.7309460053555012, "grad_norm": 0.1210460290312767, "learning_rate": 4.729479813268602e-05, "loss": 0.0272, "step": 86620 }, { "epoch": 1.731145837496503, "grad_norm": 0.10417541116476059, "learning_rate": 4.7284479140622045e-05, "loss": 0.0236, "step": 86630 }, { "epoch": 1.7313456696375045, "grad_norm": 0.14690718054771423, "learning_rate": 4.727416026456012e-05, "loss": 0.0256, "step": 86640 }, { "epoch": 1.731545501778506, "grad_norm": 0.11831032484769821, "learning_rate": 4.726384150494108e-05, "loss": 0.023, "step": 86650 }, { "epoch": 1.7317453339195077, "grad_norm": 0.11607551574707031, "learning_rate": 4.725352286220568e-05, "loss": 0.03, "step": 86660 }, { "epoch": 1.7319451660605092, "grad_norm": 0.11101231724023819, "learning_rate": 4.7243204336794746e-05, "loss": 0.026, "step": 86670 }, { "epoch": 1.7321449982015107, "grad_norm": 0.11341546475887299, "learning_rate": 4.723288592914907e-05, "loss": 0.0343, "step": 86680 }, { "epoch": 1.7323448303425124, "grad_norm": 0.14507228136062622, "learning_rate": 4.7222567639709416e-05, "loss": 0.0322, "step": 86690 }, { "epoch": 1.7325446624835137, "grad_norm": 0.13221405446529388, "learning_rate": 4.7212249468916575e-05, "loss": 0.029, "step": 86700 }, { "epoch": 1.7327444946245154, "grad_norm": 0.11579325050115585, "learning_rate": 4.72019314172113e-05, "loss": 0.0291, "step": 86710 }, { "epoch": 1.7329443267655171, "grad_norm": 0.13721662759780884, "learning_rate": 4.719161348503439e-05, "loss": 0.0252, "step": 86720 }, { "epoch": 1.7331441589065184, "grad_norm": 0.09761233627796173, "learning_rate": 4.718129567282659e-05, "loss": 0.0267, "step": 86730 }, { "epoch": 1.7333439910475201, "grad_norm": 0.14499934017658234, "learning_rate": 4.717097798102867e-05, "loss": 0.0256, "step": 86740 }, { "epoch": 1.7335438231885216, "grad_norm": 0.13299381732940674, "learning_rate": 4.7160660410081346e-05, "loss": 0.0249, "step": 86750 }, { "epoch": 1.7337436553295231, "grad_norm": 0.11841364949941635, "learning_rate": 4.715034296042541e-05, "loss": 0.0236, "step": 86760 }, { "epoch": 1.7339434874705248, "grad_norm": 0.17519234120845795, "learning_rate": 4.714002563250159e-05, "loss": 0.0408, "step": 86770 }, { "epoch": 1.7341433196115263, "grad_norm": 0.14443179965019226, "learning_rate": 4.712970842675062e-05, "loss": 0.0268, "step": 86780 }, { "epoch": 1.7343431517525278, "grad_norm": 0.21130630373954773, "learning_rate": 4.7119391343613244e-05, "loss": 0.0287, "step": 86790 }, { "epoch": 1.7345429838935296, "grad_norm": 0.13476425409317017, "learning_rate": 4.710907438353016e-05, "loss": 0.0301, "step": 86800 }, { "epoch": 1.7347428160345308, "grad_norm": 0.11485666781663895, "learning_rate": 4.709875754694211e-05, "loss": 0.0227, "step": 86810 }, { "epoch": 1.7349426481755326, "grad_norm": 0.1265757530927658, "learning_rate": 4.7088440834289815e-05, "loss": 0.0253, "step": 86820 }, { "epoch": 1.735142480316534, "grad_norm": 0.15318618714809418, "learning_rate": 4.707812424601397e-05, "loss": 0.0279, "step": 86830 }, { "epoch": 1.7353423124575356, "grad_norm": 0.11444048583507538, "learning_rate": 4.7067807782555284e-05, "loss": 0.032, "step": 86840 }, { "epoch": 1.7355421445985373, "grad_norm": 0.11303739249706268, "learning_rate": 4.7057491444354484e-05, "loss": 0.0258, "step": 86850 }, { "epoch": 1.7357419767395388, "grad_norm": 0.10703360289335251, "learning_rate": 4.7047175231852234e-05, "loss": 0.0328, "step": 86860 }, { "epoch": 1.7359418088805403, "grad_norm": 0.1306387037038803, "learning_rate": 4.7036859145489243e-05, "loss": 0.0306, "step": 86870 }, { "epoch": 1.736141641021542, "grad_norm": 0.1335705816745758, "learning_rate": 4.702654318570617e-05, "loss": 0.0278, "step": 86880 }, { "epoch": 1.7363414731625435, "grad_norm": 0.13645796477794647, "learning_rate": 4.70162273529437e-05, "loss": 0.0319, "step": 86890 }, { "epoch": 1.736541305303545, "grad_norm": 0.09846644848585129, "learning_rate": 4.700591164764255e-05, "loss": 0.0304, "step": 86900 }, { "epoch": 1.7367411374445467, "grad_norm": 0.12396273761987686, "learning_rate": 4.699559607024333e-05, "loss": 0.0268, "step": 86910 }, { "epoch": 1.736940969585548, "grad_norm": 0.1284368932247162, "learning_rate": 4.698528062118674e-05, "loss": 0.0325, "step": 86920 }, { "epoch": 1.7371408017265497, "grad_norm": 0.268484890460968, "learning_rate": 4.69749653009134e-05, "loss": 0.0301, "step": 86930 }, { "epoch": 1.7373406338675512, "grad_norm": 0.15807665884494781, "learning_rate": 4.6964650109864006e-05, "loss": 0.031, "step": 86940 }, { "epoch": 1.7375404660085527, "grad_norm": 0.15444006025791168, "learning_rate": 4.6954335048479186e-05, "loss": 0.0267, "step": 86950 }, { "epoch": 1.7377402981495544, "grad_norm": 0.1317872405052185, "learning_rate": 4.694402011719957e-05, "loss": 0.0257, "step": 86960 }, { "epoch": 1.737940130290556, "grad_norm": 0.14516927301883698, "learning_rate": 4.693370531646581e-05, "loss": 0.0381, "step": 86970 }, { "epoch": 1.7381399624315574, "grad_norm": 0.11363445222377777, "learning_rate": 4.69233906467185e-05, "loss": 0.0333, "step": 86980 }, { "epoch": 1.7383397945725592, "grad_norm": 0.26614874601364136, "learning_rate": 4.691307610839831e-05, "loss": 0.0236, "step": 86990 }, { "epoch": 1.7385396267135607, "grad_norm": 0.11535194516181946, "learning_rate": 4.690276170194584e-05, "loss": 0.0305, "step": 87000 }, { "epoch": 1.7387394588545622, "grad_norm": 0.13487125933170319, "learning_rate": 4.689244742780169e-05, "loss": 0.0269, "step": 87010 }, { "epoch": 1.7389392909955639, "grad_norm": 0.12767896056175232, "learning_rate": 4.688213328640647e-05, "loss": 0.028, "step": 87020 }, { "epoch": 1.7391391231365652, "grad_norm": 0.1211358979344368, "learning_rate": 4.687181927820081e-05, "loss": 0.0333, "step": 87030 }, { "epoch": 1.7393389552775669, "grad_norm": 0.15400199592113495, "learning_rate": 4.686150540362527e-05, "loss": 0.0285, "step": 87040 }, { "epoch": 1.7395387874185684, "grad_norm": 0.1135825365781784, "learning_rate": 4.685119166312046e-05, "loss": 0.0277, "step": 87050 }, { "epoch": 1.7397386195595699, "grad_norm": 0.09286938607692719, "learning_rate": 4.684087805712696e-05, "loss": 0.0276, "step": 87060 }, { "epoch": 1.7399384517005716, "grad_norm": 0.1028568223118782, "learning_rate": 4.683056458608533e-05, "loss": 0.0312, "step": 87070 }, { "epoch": 1.740138283841573, "grad_norm": 0.13483914732933044, "learning_rate": 4.682025125043616e-05, "loss": 0.0264, "step": 87080 }, { "epoch": 1.7403381159825746, "grad_norm": 0.10945086181163788, "learning_rate": 4.680993805062004e-05, "loss": 0.0245, "step": 87090 }, { "epoch": 1.7405379481235763, "grad_norm": 0.135044664144516, "learning_rate": 4.679962498707748e-05, "loss": 0.0284, "step": 87100 }, { "epoch": 1.7407377802645776, "grad_norm": 0.1354655921459198, "learning_rate": 4.678931206024906e-05, "loss": 0.0269, "step": 87110 }, { "epoch": 1.7409376124055793, "grad_norm": 0.15320910513401031, "learning_rate": 4.6778999270575355e-05, "loss": 0.0359, "step": 87120 }, { "epoch": 1.741137444546581, "grad_norm": 0.14792205393314362, "learning_rate": 4.676868661849687e-05, "loss": 0.0226, "step": 87130 }, { "epoch": 1.7413372766875823, "grad_norm": 0.11079208552837372, "learning_rate": 4.675837410445417e-05, "loss": 0.0236, "step": 87140 }, { "epoch": 1.741537108828584, "grad_norm": 0.11533304303884506, "learning_rate": 4.674806172888776e-05, "loss": 0.0214, "step": 87150 }, { "epoch": 1.7417369409695855, "grad_norm": 0.1306879073381424, "learning_rate": 4.6737749492238166e-05, "loss": 0.0307, "step": 87160 }, { "epoch": 1.741936773110587, "grad_norm": 0.17681913077831268, "learning_rate": 4.672743739494594e-05, "loss": 0.0343, "step": 87170 }, { "epoch": 1.7421366052515888, "grad_norm": 0.12339874356985092, "learning_rate": 4.671712543745157e-05, "loss": 0.0234, "step": 87180 }, { "epoch": 1.7423364373925903, "grad_norm": 0.19188342988491058, "learning_rate": 4.6706813620195586e-05, "loss": 0.0337, "step": 87190 }, { "epoch": 1.7425362695335918, "grad_norm": 0.08437848091125488, "learning_rate": 4.6696501943618446e-05, "loss": 0.03, "step": 87200 }, { "epoch": 1.7427361016745935, "grad_norm": 0.1274702548980713, "learning_rate": 4.668619040816069e-05, "loss": 0.0345, "step": 87210 }, { "epoch": 1.7429359338155948, "grad_norm": 0.15941940248012543, "learning_rate": 4.667587901426279e-05, "loss": 0.0313, "step": 87220 }, { "epoch": 1.7431357659565965, "grad_norm": 0.10691273212432861, "learning_rate": 4.6665567762365234e-05, "loss": 0.0268, "step": 87230 }, { "epoch": 1.743335598097598, "grad_norm": 0.14264196157455444, "learning_rate": 4.6655256652908507e-05, "loss": 0.0303, "step": 87240 }, { "epoch": 1.7435354302385995, "grad_norm": 0.11510687321424484, "learning_rate": 4.6644945686333044e-05, "loss": 0.0245, "step": 87250 }, { "epoch": 1.7437352623796012, "grad_norm": 0.11574311554431915, "learning_rate": 4.663463486307935e-05, "loss": 0.0319, "step": 87260 }, { "epoch": 1.7439350945206027, "grad_norm": 0.09919941425323486, "learning_rate": 4.662432418358789e-05, "loss": 0.0296, "step": 87270 }, { "epoch": 1.7441349266616042, "grad_norm": 0.14243696630001068, "learning_rate": 4.661401364829908e-05, "loss": 0.0299, "step": 87280 }, { "epoch": 1.744334758802606, "grad_norm": 0.09441355615854263, "learning_rate": 4.6603703257653386e-05, "loss": 0.0235, "step": 87290 }, { "epoch": 1.7445345909436074, "grad_norm": 0.149446502327919, "learning_rate": 4.659339301209126e-05, "loss": 0.0378, "step": 87300 }, { "epoch": 1.744734423084609, "grad_norm": 0.13114912807941437, "learning_rate": 4.658308291205312e-05, "loss": 0.0277, "step": 87310 }, { "epoch": 1.7449342552256106, "grad_norm": 0.1264735609292984, "learning_rate": 4.6572772957979416e-05, "loss": 0.0303, "step": 87320 }, { "epoch": 1.745134087366612, "grad_norm": 0.16158807277679443, "learning_rate": 4.656246315031054e-05, "loss": 0.0238, "step": 87330 }, { "epoch": 1.7453339195076136, "grad_norm": 0.12747594714164734, "learning_rate": 4.655215348948692e-05, "loss": 0.0246, "step": 87340 }, { "epoch": 1.7455337516486151, "grad_norm": 0.1347423791885376, "learning_rate": 4.6541843975948976e-05, "loss": 0.0322, "step": 87350 }, { "epoch": 1.7457335837896166, "grad_norm": 0.12730000913143158, "learning_rate": 4.65315346101371e-05, "loss": 0.0327, "step": 87360 }, { "epoch": 1.7459334159306183, "grad_norm": 0.1390756219625473, "learning_rate": 4.6521225392491716e-05, "loss": 0.0368, "step": 87370 }, { "epoch": 1.7461332480716198, "grad_norm": 0.1116681694984436, "learning_rate": 4.651091632345316e-05, "loss": 0.0258, "step": 87380 }, { "epoch": 1.7463330802126213, "grad_norm": 0.1266268640756607, "learning_rate": 4.650060740346186e-05, "loss": 0.0258, "step": 87390 }, { "epoch": 1.746532912353623, "grad_norm": 0.1596807837486267, "learning_rate": 4.6490298632958196e-05, "loss": 0.0365, "step": 87400 }, { "epoch": 1.7467327444946246, "grad_norm": 0.14600799977779388, "learning_rate": 4.647999001238251e-05, "loss": 0.0277, "step": 87410 }, { "epoch": 1.746932576635626, "grad_norm": 0.14225627481937408, "learning_rate": 4.6469681542175206e-05, "loss": 0.0313, "step": 87420 }, { "epoch": 1.7471324087766278, "grad_norm": 0.1196644976735115, "learning_rate": 4.645937322277658e-05, "loss": 0.0259, "step": 87430 }, { "epoch": 1.747332240917629, "grad_norm": 0.15613597631454468, "learning_rate": 4.644906505462706e-05, "loss": 0.0365, "step": 87440 }, { "epoch": 1.7475320730586308, "grad_norm": 0.11079802364110947, "learning_rate": 4.6438757038166955e-05, "loss": 0.0364, "step": 87450 }, { "epoch": 1.7477319051996323, "grad_norm": 0.10925239324569702, "learning_rate": 4.6428449173836604e-05, "loss": 0.0284, "step": 87460 }, { "epoch": 1.7479317373406338, "grad_norm": 0.09553607553243637, "learning_rate": 4.641814146207633e-05, "loss": 0.037, "step": 87470 }, { "epoch": 1.7481315694816355, "grad_norm": 0.11002694815397263, "learning_rate": 4.640783390332648e-05, "loss": 0.0265, "step": 87480 }, { "epoch": 1.748331401622637, "grad_norm": 0.1402100920677185, "learning_rate": 4.6397526498027375e-05, "loss": 0.0341, "step": 87490 }, { "epoch": 1.7485312337636385, "grad_norm": 0.1018519252538681, "learning_rate": 4.6387219246619315e-05, "loss": 0.0255, "step": 87500 }, { "epoch": 1.7487310659046402, "grad_norm": 0.11991086602210999, "learning_rate": 4.63769121495426e-05, "loss": 0.0281, "step": 87510 }, { "epoch": 1.7489308980456415, "grad_norm": 0.13949918746948242, "learning_rate": 4.6366605207237566e-05, "loss": 0.0253, "step": 87520 }, { "epoch": 1.7491307301866432, "grad_norm": 0.14464038610458374, "learning_rate": 4.6356298420144466e-05, "loss": 0.0271, "step": 87530 }, { "epoch": 1.749330562327645, "grad_norm": 0.10338117182254791, "learning_rate": 4.634599178870362e-05, "loss": 0.0324, "step": 87540 }, { "epoch": 1.7495303944686462, "grad_norm": 0.14447833597660065, "learning_rate": 4.6335685313355275e-05, "loss": 0.0236, "step": 87550 }, { "epoch": 1.749730226609648, "grad_norm": 0.12578298151493073, "learning_rate": 4.632537899453971e-05, "loss": 0.0365, "step": 87560 }, { "epoch": 1.7499300587506494, "grad_norm": 0.1491059809923172, "learning_rate": 4.6315072832697236e-05, "loss": 0.0266, "step": 87570 }, { "epoch": 1.750129890891651, "grad_norm": 0.14943234622478485, "learning_rate": 4.630476682826806e-05, "loss": 0.0205, "step": 87580 }, { "epoch": 1.7503297230326527, "grad_norm": 0.10830210894346237, "learning_rate": 4.6294460981692476e-05, "loss": 0.0325, "step": 87590 }, { "epoch": 1.7505295551736542, "grad_norm": 0.11011906713247299, "learning_rate": 4.628415529341069e-05, "loss": 0.0263, "step": 87600 }, { "epoch": 1.7507293873146557, "grad_norm": 0.14227578043937683, "learning_rate": 4.627384976386297e-05, "loss": 0.0289, "step": 87610 }, { "epoch": 1.7509292194556574, "grad_norm": 0.1272665411233902, "learning_rate": 4.626354439348956e-05, "loss": 0.0241, "step": 87620 }, { "epoch": 1.7511290515966587, "grad_norm": 0.13812604546546936, "learning_rate": 4.625323918273065e-05, "loss": 0.0305, "step": 87630 }, { "epoch": 1.7513288837376604, "grad_norm": 0.09862114489078522, "learning_rate": 4.624293413202649e-05, "loss": 0.0314, "step": 87640 }, { "epoch": 1.7515287158786619, "grad_norm": 0.11841429769992828, "learning_rate": 4.623262924181727e-05, "loss": 0.027, "step": 87650 }, { "epoch": 1.7517285480196634, "grad_norm": 0.15458151698112488, "learning_rate": 4.62223245125432e-05, "loss": 0.0237, "step": 87660 }, { "epoch": 1.751928380160665, "grad_norm": 0.12614302337169647, "learning_rate": 4.621201994464451e-05, "loss": 0.0262, "step": 87670 }, { "epoch": 1.7521282123016666, "grad_norm": 0.10206060856580734, "learning_rate": 4.6201715538561366e-05, "loss": 0.033, "step": 87680 }, { "epoch": 1.752328044442668, "grad_norm": 0.1358693689107895, "learning_rate": 4.619141129473393e-05, "loss": 0.0339, "step": 87690 }, { "epoch": 1.7525278765836698, "grad_norm": 0.10970605164766312, "learning_rate": 4.618110721360244e-05, "loss": 0.0331, "step": 87700 }, { "epoch": 1.7527277087246713, "grad_norm": 0.30710628628730774, "learning_rate": 4.617080329560702e-05, "loss": 0.0351, "step": 87710 }, { "epoch": 1.7529275408656728, "grad_norm": 0.11867941170930862, "learning_rate": 4.616049954118786e-05, "loss": 0.0357, "step": 87720 }, { "epoch": 1.7531273730066745, "grad_norm": 0.11913306266069412, "learning_rate": 4.6150195950785095e-05, "loss": 0.0339, "step": 87730 }, { "epoch": 1.7533272051476758, "grad_norm": 0.1376631259918213, "learning_rate": 4.613989252483888e-05, "loss": 0.0299, "step": 87740 }, { "epoch": 1.7535270372886775, "grad_norm": 0.10207506269216537, "learning_rate": 4.612958926378938e-05, "loss": 0.0253, "step": 87750 }, { "epoch": 1.753726869429679, "grad_norm": 0.12372930347919464, "learning_rate": 4.6119286168076706e-05, "loss": 0.0263, "step": 87760 }, { "epoch": 1.7539267015706805, "grad_norm": 0.10235287249088287, "learning_rate": 4.610898323814101e-05, "loss": 0.0286, "step": 87770 }, { "epoch": 1.7541265337116823, "grad_norm": 0.09328021109104156, "learning_rate": 4.6098680474422386e-05, "loss": 0.0249, "step": 87780 }, { "epoch": 1.7543263658526838, "grad_norm": 0.12537415325641632, "learning_rate": 4.608837787736096e-05, "loss": 0.0326, "step": 87790 }, { "epoch": 1.7545261979936853, "grad_norm": 0.13240313529968262, "learning_rate": 4.607807544739688e-05, "loss": 0.0265, "step": 87800 }, { "epoch": 1.754726030134687, "grad_norm": 0.1351890116930008, "learning_rate": 4.606777318497019e-05, "loss": 0.0205, "step": 87810 }, { "epoch": 1.7549258622756885, "grad_norm": 0.11292880028486252, "learning_rate": 4.6057471090521006e-05, "loss": 0.0269, "step": 87820 }, { "epoch": 1.75512569441669, "grad_norm": 0.22173359990119934, "learning_rate": 4.60471691644894e-05, "loss": 0.0308, "step": 87830 }, { "epoch": 1.7553255265576917, "grad_norm": 0.13933317363262177, "learning_rate": 4.603686740731549e-05, "loss": 0.0311, "step": 87840 }, { "epoch": 1.755525358698693, "grad_norm": 0.11373941600322723, "learning_rate": 4.6026565819439306e-05, "loss": 0.0285, "step": 87850 }, { "epoch": 1.7557251908396947, "grad_norm": 0.11435351520776749, "learning_rate": 4.6016264401300944e-05, "loss": 0.0315, "step": 87860 }, { "epoch": 1.7559250229806962, "grad_norm": 0.14166194200515747, "learning_rate": 4.600596315334044e-05, "loss": 0.0305, "step": 87870 }, { "epoch": 1.7561248551216977, "grad_norm": 0.0839901939034462, "learning_rate": 4.599566207599785e-05, "loss": 0.0291, "step": 87880 }, { "epoch": 1.7563246872626994, "grad_norm": 0.12917682528495789, "learning_rate": 4.598536116971324e-05, "loss": 0.0343, "step": 87890 }, { "epoch": 1.756524519403701, "grad_norm": 0.11197825521230698, "learning_rate": 4.59750604349266e-05, "loss": 0.0345, "step": 87900 }, { "epoch": 1.7567243515447024, "grad_norm": 0.11504685878753662, "learning_rate": 4.596475987207801e-05, "loss": 0.0264, "step": 87910 }, { "epoch": 1.7569241836857041, "grad_norm": 0.1465335488319397, "learning_rate": 4.595445948160744e-05, "loss": 0.0255, "step": 87920 }, { "epoch": 1.7571240158267054, "grad_norm": 0.13278159499168396, "learning_rate": 4.5944159263954935e-05, "loss": 0.0289, "step": 87930 }, { "epoch": 1.7573238479677071, "grad_norm": 0.14915819466114044, "learning_rate": 4.5933859219560507e-05, "loss": 0.0296, "step": 87940 }, { "epoch": 1.7575236801087086, "grad_norm": 0.13560567796230316, "learning_rate": 4.5923559348864136e-05, "loss": 0.0289, "step": 87950 }, { "epoch": 1.7577235122497101, "grad_norm": 0.12802980840206146, "learning_rate": 4.591325965230581e-05, "loss": 0.028, "step": 87960 }, { "epoch": 1.7579233443907119, "grad_norm": 0.12330962717533112, "learning_rate": 4.590296013032555e-05, "loss": 0.0421, "step": 87970 }, { "epoch": 1.7581231765317133, "grad_norm": 0.12194294482469559, "learning_rate": 4.589266078336329e-05, "loss": 0.0241, "step": 87980 }, { "epoch": 1.7583230086727148, "grad_norm": 0.12234470248222351, "learning_rate": 4.5882361611859026e-05, "loss": 0.0211, "step": 87990 }, { "epoch": 1.7585228408137166, "grad_norm": 0.1177382692694664, "learning_rate": 4.5872062616252695e-05, "loss": 0.032, "step": 88000 }, { "epoch": 1.758722672954718, "grad_norm": 0.08597826212644577, "learning_rate": 4.5861763796984265e-05, "loss": 0.0211, "step": 88010 }, { "epoch": 1.7589225050957196, "grad_norm": 0.10135753452777863, "learning_rate": 4.58514651544937e-05, "loss": 0.024, "step": 88020 }, { "epoch": 1.7591223372367213, "grad_norm": 0.1271415799856186, "learning_rate": 4.5841166689220915e-05, "loss": 0.031, "step": 88030 }, { "epoch": 1.7593221693777226, "grad_norm": 0.12221497297286987, "learning_rate": 4.583086840160585e-05, "loss": 0.0354, "step": 88040 }, { "epoch": 1.7595220015187243, "grad_norm": 0.13176684081554413, "learning_rate": 4.582057029208841e-05, "loss": 0.0263, "step": 88050 }, { "epoch": 1.7597218336597258, "grad_norm": 0.10541830211877823, "learning_rate": 4.581027236110855e-05, "loss": 0.0318, "step": 88060 }, { "epoch": 1.7599216658007273, "grad_norm": 0.09366851300001144, "learning_rate": 4.579997460910617e-05, "loss": 0.0257, "step": 88070 }, { "epoch": 1.760121497941729, "grad_norm": 0.16024819016456604, "learning_rate": 4.578967703652114e-05, "loss": 0.0368, "step": 88080 }, { "epoch": 1.7603213300827305, "grad_norm": 0.14228665828704834, "learning_rate": 4.577937964379338e-05, "loss": 0.0295, "step": 88090 }, { "epoch": 1.760521162223732, "grad_norm": 0.1134485974907875, "learning_rate": 4.576908243136275e-05, "loss": 0.0341, "step": 88100 }, { "epoch": 1.7607209943647337, "grad_norm": 0.14694929122924805, "learning_rate": 4.5758785399669154e-05, "loss": 0.0301, "step": 88110 }, { "epoch": 1.7609208265057352, "grad_norm": 0.14419527351856232, "learning_rate": 4.574848854915247e-05, "loss": 0.0243, "step": 88120 }, { "epoch": 1.7611206586467367, "grad_norm": 0.11233272403478622, "learning_rate": 4.5738191880252526e-05, "loss": 0.0306, "step": 88130 }, { "epoch": 1.7613204907877384, "grad_norm": 0.14533859491348267, "learning_rate": 4.5727895393409195e-05, "loss": 0.0286, "step": 88140 }, { "epoch": 1.7615203229287397, "grad_norm": 0.12258774787187576, "learning_rate": 4.571759908906234e-05, "loss": 0.0264, "step": 88150 }, { "epoch": 1.7617201550697414, "grad_norm": 0.1601257026195526, "learning_rate": 4.5707302967651775e-05, "loss": 0.0269, "step": 88160 }, { "epoch": 1.761919987210743, "grad_norm": 0.13465042412281036, "learning_rate": 4.569700702961735e-05, "loss": 0.031, "step": 88170 }, { "epoch": 1.7621198193517444, "grad_norm": 0.12448496371507645, "learning_rate": 4.568671127539888e-05, "loss": 0.0292, "step": 88180 }, { "epoch": 1.7623196514927462, "grad_norm": 0.1280883550643921, "learning_rate": 4.567641570543616e-05, "loss": 0.027, "step": 88190 }, { "epoch": 1.7625194836337477, "grad_norm": 0.11663824319839478, "learning_rate": 4.566612032016905e-05, "loss": 0.0294, "step": 88200 }, { "epoch": 1.7627193157747492, "grad_norm": 0.10655885934829712, "learning_rate": 4.5655825120037294e-05, "loss": 0.0221, "step": 88210 }, { "epoch": 1.7629191479157509, "grad_norm": 0.11755818128585815, "learning_rate": 4.5645530105480726e-05, "loss": 0.026, "step": 88220 }, { "epoch": 1.7631189800567524, "grad_norm": 0.11989406496286392, "learning_rate": 4.5635235276939084e-05, "loss": 0.0272, "step": 88230 }, { "epoch": 1.7633188121977539, "grad_norm": 0.11575566232204437, "learning_rate": 4.562494063485219e-05, "loss": 0.0216, "step": 88240 }, { "epoch": 1.7635186443387556, "grad_norm": 0.15721020102500916, "learning_rate": 4.5614646179659796e-05, "loss": 0.0253, "step": 88250 }, { "epoch": 1.7637184764797569, "grad_norm": 0.09560763090848923, "learning_rate": 4.560435191180167e-05, "loss": 0.0307, "step": 88260 }, { "epoch": 1.7639183086207586, "grad_norm": 0.11260923743247986, "learning_rate": 4.559405783171755e-05, "loss": 0.0269, "step": 88270 }, { "epoch": 1.76411814076176, "grad_norm": 0.09433087706565857, "learning_rate": 4.5583763939847166e-05, "loss": 0.027, "step": 88280 }, { "epoch": 1.7643179729027616, "grad_norm": 0.12272656708955765, "learning_rate": 4.55734702366303e-05, "loss": 0.0338, "step": 88290 }, { "epoch": 1.7645178050437633, "grad_norm": 0.12497609853744507, "learning_rate": 4.556317672250664e-05, "loss": 0.0236, "step": 88300 }, { "epoch": 1.7647176371847648, "grad_norm": 0.09897273033857346, "learning_rate": 4.555288339791593e-05, "loss": 0.0267, "step": 88310 }, { "epoch": 1.7649174693257663, "grad_norm": 0.13290873169898987, "learning_rate": 4.554259026329785e-05, "loss": 0.0257, "step": 88320 }, { "epoch": 1.765117301466768, "grad_norm": 0.09731565415859222, "learning_rate": 4.553229731909214e-05, "loss": 0.0278, "step": 88330 }, { "epoch": 1.7653171336077693, "grad_norm": 0.11120029538869858, "learning_rate": 4.552200456573849e-05, "loss": 0.0294, "step": 88340 }, { "epoch": 1.765516965748771, "grad_norm": 0.1426791548728943, "learning_rate": 4.551171200367657e-05, "loss": 0.025, "step": 88350 }, { "epoch": 1.7657167978897725, "grad_norm": 0.13473382592201233, "learning_rate": 4.5501419633346074e-05, "loss": 0.0311, "step": 88360 }, { "epoch": 1.765916630030774, "grad_norm": 0.12491922080516815, "learning_rate": 4.549112745518665e-05, "loss": 0.0238, "step": 88370 }, { "epoch": 1.7661164621717758, "grad_norm": 0.12138131260871887, "learning_rate": 4.548083546963798e-05, "loss": 0.0326, "step": 88380 }, { "epoch": 1.7663162943127773, "grad_norm": 0.13115061819553375, "learning_rate": 4.547054367713974e-05, "loss": 0.0213, "step": 88390 }, { "epoch": 1.7665161264537788, "grad_norm": 0.12172222882509232, "learning_rate": 4.5460252078131537e-05, "loss": 0.026, "step": 88400 }, { "epoch": 1.7667159585947805, "grad_norm": 0.12906430661678314, "learning_rate": 4.5449960673053014e-05, "loss": 0.0291, "step": 88410 }, { "epoch": 1.766915790735782, "grad_norm": 0.14320480823516846, "learning_rate": 4.5439669462343834e-05, "loss": 0.033, "step": 88420 }, { "epoch": 1.7671156228767835, "grad_norm": 0.10387525707483292, "learning_rate": 4.542937844644359e-05, "loss": 0.0335, "step": 88430 }, { "epoch": 1.7673154550177852, "grad_norm": 0.10680907219648361, "learning_rate": 4.5419087625791904e-05, "loss": 0.031, "step": 88440 }, { "epoch": 1.7675152871587865, "grad_norm": 0.17937041819095612, "learning_rate": 4.540879700082838e-05, "loss": 0.0257, "step": 88450 }, { "epoch": 1.7677151192997882, "grad_norm": 0.10934338718652725, "learning_rate": 4.539850657199259e-05, "loss": 0.0215, "step": 88460 }, { "epoch": 1.7679149514407897, "grad_norm": 0.08927162736654282, "learning_rate": 4.538821633972418e-05, "loss": 0.028, "step": 88470 }, { "epoch": 1.7681147835817912, "grad_norm": 0.11549254506826401, "learning_rate": 4.537792630446268e-05, "loss": 0.0269, "step": 88480 }, { "epoch": 1.768314615722793, "grad_norm": 0.1140807494521141, "learning_rate": 4.536763646664769e-05, "loss": 0.0268, "step": 88490 }, { "epoch": 1.7685144478637944, "grad_norm": 0.10861809551715851, "learning_rate": 4.535734682671873e-05, "loss": 0.0337, "step": 88500 }, { "epoch": 1.768714280004796, "grad_norm": 0.11157047748565674, "learning_rate": 4.53470573851154e-05, "loss": 0.0362, "step": 88510 }, { "epoch": 1.7689141121457976, "grad_norm": 0.1192188560962677, "learning_rate": 4.5336768142277244e-05, "loss": 0.0301, "step": 88520 }, { "epoch": 1.7691139442867991, "grad_norm": 0.11849337071180344, "learning_rate": 4.532647909864378e-05, "loss": 0.0252, "step": 88530 }, { "epoch": 1.7693137764278006, "grad_norm": 0.12730024755001068, "learning_rate": 4.531619025465455e-05, "loss": 0.0254, "step": 88540 }, { "epoch": 1.7695136085688024, "grad_norm": 0.14893080294132233, "learning_rate": 4.5305901610749046e-05, "loss": 0.0309, "step": 88550 }, { "epoch": 1.7697134407098036, "grad_norm": 0.11766532808542252, "learning_rate": 4.5295613167366815e-05, "loss": 0.0313, "step": 88560 }, { "epoch": 1.7699132728508054, "grad_norm": 0.13683487474918365, "learning_rate": 4.528532492494735e-05, "loss": 0.0324, "step": 88570 }, { "epoch": 1.7701131049918069, "grad_norm": 0.11752991378307343, "learning_rate": 4.527503688393014e-05, "loss": 0.0229, "step": 88580 }, { "epoch": 1.7703129371328084, "grad_norm": 0.1446552574634552, "learning_rate": 4.526474904475466e-05, "loss": 0.0237, "step": 88590 }, { "epoch": 1.77051276927381, "grad_norm": 0.13970844447612762, "learning_rate": 4.525446140786043e-05, "loss": 0.0359, "step": 88600 }, { "epoch": 1.7707126014148116, "grad_norm": 0.13559797406196594, "learning_rate": 4.5244173973686866e-05, "loss": 0.0345, "step": 88610 }, { "epoch": 1.770912433555813, "grad_norm": 0.13972054421901703, "learning_rate": 4.5233886742673476e-05, "loss": 0.0366, "step": 88620 }, { "epoch": 1.7711122656968148, "grad_norm": 0.1657765954732895, "learning_rate": 4.5223599715259675e-05, "loss": 0.0272, "step": 88630 }, { "epoch": 1.771312097837816, "grad_norm": 0.1177096962928772, "learning_rate": 4.521331289188491e-05, "loss": 0.0279, "step": 88640 }, { "epoch": 1.7715119299788178, "grad_norm": 0.1160285472869873, "learning_rate": 4.520302627298863e-05, "loss": 0.0263, "step": 88650 }, { "epoch": 1.7717117621198195, "grad_norm": 0.14919893443584442, "learning_rate": 4.519273985901027e-05, "loss": 0.0308, "step": 88660 }, { "epoch": 1.7719115942608208, "grad_norm": 0.16502387821674347, "learning_rate": 4.518245365038922e-05, "loss": 0.0263, "step": 88670 }, { "epoch": 1.7721114264018225, "grad_norm": 0.1050289049744606, "learning_rate": 4.517216764756489e-05, "loss": 0.0339, "step": 88680 }, { "epoch": 1.772311258542824, "grad_norm": 0.11587341129779816, "learning_rate": 4.516188185097671e-05, "loss": 0.0255, "step": 88690 }, { "epoch": 1.7725110906838255, "grad_norm": 0.1105901375412941, "learning_rate": 4.5151596261064044e-05, "loss": 0.0243, "step": 88700 }, { "epoch": 1.7727109228248272, "grad_norm": 0.12590709328651428, "learning_rate": 4.5141310878266286e-05, "loss": 0.0336, "step": 88710 }, { "epoch": 1.7729107549658287, "grad_norm": 0.1178727000951767, "learning_rate": 4.513102570302278e-05, "loss": 0.0317, "step": 88720 }, { "epoch": 1.7731105871068302, "grad_norm": 0.10387185961008072, "learning_rate": 4.512074073577291e-05, "loss": 0.0212, "step": 88730 }, { "epoch": 1.773310419247832, "grad_norm": 0.1047796756029129, "learning_rate": 4.511045597695606e-05, "loss": 0.0259, "step": 88740 }, { "epoch": 1.7735102513888332, "grad_norm": 0.09593769162893295, "learning_rate": 4.5100171427011534e-05, "loss": 0.0334, "step": 88750 }, { "epoch": 1.773710083529835, "grad_norm": 0.11960854381322861, "learning_rate": 4.508988708637869e-05, "loss": 0.0216, "step": 88760 }, { "epoch": 1.7739099156708364, "grad_norm": 0.13484078645706177, "learning_rate": 4.507960295549683e-05, "loss": 0.0246, "step": 88770 }, { "epoch": 1.774109747811838, "grad_norm": 0.1283811777830124, "learning_rate": 4.5069319034805306e-05, "loss": 0.0295, "step": 88780 }, { "epoch": 1.7743095799528397, "grad_norm": 0.12896355986595154, "learning_rate": 4.505903532474342e-05, "loss": 0.0301, "step": 88790 }, { "epoch": 1.7745094120938412, "grad_norm": 0.11427924036979675, "learning_rate": 4.5048751825750466e-05, "loss": 0.0331, "step": 88800 }, { "epoch": 1.7747092442348427, "grad_norm": 0.12989410758018494, "learning_rate": 4.503846853826572e-05, "loss": 0.0296, "step": 88810 }, { "epoch": 1.7749090763758444, "grad_norm": 0.12695513665676117, "learning_rate": 4.5028185462728505e-05, "loss": 0.0293, "step": 88820 }, { "epoch": 1.7751089085168459, "grad_norm": 0.1569482535123825, "learning_rate": 4.5017902599578066e-05, "loss": 0.0453, "step": 88830 }, { "epoch": 1.7753087406578474, "grad_norm": 0.11393950134515762, "learning_rate": 4.500761994925369e-05, "loss": 0.0274, "step": 88840 }, { "epoch": 1.775508572798849, "grad_norm": 0.10761445015668869, "learning_rate": 4.49973375121946e-05, "loss": 0.032, "step": 88850 }, { "epoch": 1.7757084049398504, "grad_norm": 0.10502917319536209, "learning_rate": 4.498705528884005e-05, "loss": 0.0287, "step": 88860 }, { "epoch": 1.775908237080852, "grad_norm": 0.11520829051733017, "learning_rate": 4.497677327962931e-05, "loss": 0.021, "step": 88870 }, { "epoch": 1.7761080692218536, "grad_norm": 0.12041796743869781, "learning_rate": 4.4966491485001585e-05, "loss": 0.0283, "step": 88880 }, { "epoch": 1.776307901362855, "grad_norm": 0.11137975007295609, "learning_rate": 4.49562099053961e-05, "loss": 0.0289, "step": 88890 }, { "epoch": 1.7765077335038568, "grad_norm": 0.11484619975090027, "learning_rate": 4.494592854125205e-05, "loss": 0.0223, "step": 88900 }, { "epoch": 1.7767075656448583, "grad_norm": 0.10306432843208313, "learning_rate": 4.493564739300865e-05, "loss": 0.0255, "step": 88910 }, { "epoch": 1.7769073977858598, "grad_norm": 0.11404672265052795, "learning_rate": 4.49253664611051e-05, "loss": 0.0316, "step": 88920 }, { "epoch": 1.7771072299268615, "grad_norm": 0.1071520522236824, "learning_rate": 4.491508574598057e-05, "loss": 0.0253, "step": 88930 }, { "epoch": 1.777307062067863, "grad_norm": 0.12853194773197174, "learning_rate": 4.490480524807424e-05, "loss": 0.0306, "step": 88940 }, { "epoch": 1.7775068942088645, "grad_norm": 0.17004401981830597, "learning_rate": 4.489452496782525e-05, "loss": 0.0216, "step": 88950 }, { "epoch": 1.7777067263498663, "grad_norm": 0.15009649097919464, "learning_rate": 4.488424490567279e-05, "loss": 0.0244, "step": 88960 }, { "epoch": 1.7779065584908675, "grad_norm": 0.11702043563127518, "learning_rate": 4.487396506205599e-05, "loss": 0.0277, "step": 88970 }, { "epoch": 1.7781063906318693, "grad_norm": 0.14340706169605255, "learning_rate": 4.4863685437413975e-05, "loss": 0.03, "step": 88980 }, { "epoch": 1.7783062227728708, "grad_norm": 0.13370700180530548, "learning_rate": 4.485340603218588e-05, "loss": 0.0275, "step": 88990 }, { "epoch": 1.7785060549138723, "grad_norm": 0.1189902126789093, "learning_rate": 4.484312684681084e-05, "loss": 0.0307, "step": 89000 }, { "epoch": 1.778705887054874, "grad_norm": 0.10027078539133072, "learning_rate": 4.483284788172794e-05, "loss": 0.0244, "step": 89010 }, { "epoch": 1.7789057191958755, "grad_norm": 0.11419939994812012, "learning_rate": 4.48225691373763e-05, "loss": 0.023, "step": 89020 }, { "epoch": 1.779105551336877, "grad_norm": 0.15861807763576508, "learning_rate": 4.481229061419498e-05, "loss": 0.033, "step": 89030 }, { "epoch": 1.7793053834778787, "grad_norm": 0.15131467580795288, "learning_rate": 4.4802012312623064e-05, "loss": 0.0221, "step": 89040 }, { "epoch": 1.77950521561888, "grad_norm": 0.11855703592300415, "learning_rate": 4.4791734233099645e-05, "loss": 0.0261, "step": 89050 }, { "epoch": 1.7797050477598817, "grad_norm": 0.15448269248008728, "learning_rate": 4.478145637606379e-05, "loss": 0.0331, "step": 89060 }, { "epoch": 1.7799048799008834, "grad_norm": 0.10661409795284271, "learning_rate": 4.477117874195451e-05, "loss": 0.0276, "step": 89070 }, { "epoch": 1.7801047120418847, "grad_norm": 0.12309840321540833, "learning_rate": 4.476090133121086e-05, "loss": 0.0289, "step": 89080 }, { "epoch": 1.7803045441828864, "grad_norm": 0.1075408011674881, "learning_rate": 4.47506241442719e-05, "loss": 0.0226, "step": 89090 }, { "epoch": 1.780504376323888, "grad_norm": 0.15406426787376404, "learning_rate": 4.474034718157663e-05, "loss": 0.0384, "step": 89100 }, { "epoch": 1.7807042084648894, "grad_norm": 0.13097716867923737, "learning_rate": 4.473007044356407e-05, "loss": 0.0324, "step": 89110 }, { "epoch": 1.7809040406058911, "grad_norm": 0.1055375263094902, "learning_rate": 4.471979393067321e-05, "loss": 0.0273, "step": 89120 }, { "epoch": 1.7811038727468926, "grad_norm": 0.1329796314239502, "learning_rate": 4.470951764334304e-05, "loss": 0.0242, "step": 89130 }, { "epoch": 1.7813037048878941, "grad_norm": 0.1204831525683403, "learning_rate": 4.469924158201259e-05, "loss": 0.022, "step": 89140 }, { "epoch": 1.7815035370288959, "grad_norm": 0.12319070845842361, "learning_rate": 4.468896574712078e-05, "loss": 0.0293, "step": 89150 }, { "epoch": 1.7817033691698971, "grad_norm": 0.12020540982484818, "learning_rate": 4.467869013910661e-05, "loss": 0.033, "step": 89160 }, { "epoch": 1.7819032013108989, "grad_norm": 0.12348666787147522, "learning_rate": 4.4668414758409e-05, "loss": 0.041, "step": 89170 }, { "epoch": 1.7821030334519004, "grad_norm": 0.1069413498044014, "learning_rate": 4.465813960546693e-05, "loss": 0.0348, "step": 89180 }, { "epoch": 1.7823028655929019, "grad_norm": 0.09882807731628418, "learning_rate": 4.4647864680719334e-05, "loss": 0.0229, "step": 89190 }, { "epoch": 1.7825026977339036, "grad_norm": 0.10173460841178894, "learning_rate": 4.463758998460511e-05, "loss": 0.0241, "step": 89200 }, { "epoch": 1.782702529874905, "grad_norm": 0.09780334681272507, "learning_rate": 4.4627315517563206e-05, "loss": 0.0265, "step": 89210 }, { "epoch": 1.7829023620159066, "grad_norm": 0.15823382139205933, "learning_rate": 4.4617041280032496e-05, "loss": 0.0284, "step": 89220 }, { "epoch": 1.7831021941569083, "grad_norm": 0.11486956477165222, "learning_rate": 4.46067672724519e-05, "loss": 0.0297, "step": 89230 }, { "epoch": 1.7833020262979098, "grad_norm": 0.09958618134260178, "learning_rate": 4.4596493495260306e-05, "loss": 0.0285, "step": 89240 }, { "epoch": 1.7835018584389113, "grad_norm": 0.1125861331820488, "learning_rate": 4.4586219948896576e-05, "loss": 0.0274, "step": 89250 }, { "epoch": 1.783701690579913, "grad_norm": 0.140193909406662, "learning_rate": 4.457594663379957e-05, "loss": 0.0228, "step": 89260 }, { "epoch": 1.7839015227209143, "grad_norm": 0.10049701482057571, "learning_rate": 4.456567355040819e-05, "loss": 0.027, "step": 89270 }, { "epoch": 1.784101354861916, "grad_norm": 0.124339759349823, "learning_rate": 4.455540069916124e-05, "loss": 0.0325, "step": 89280 }, { "epoch": 1.7843011870029175, "grad_norm": 0.19168882071971893, "learning_rate": 4.454512808049757e-05, "loss": 0.0312, "step": 89290 }, { "epoch": 1.784501019143919, "grad_norm": 0.11254201829433441, "learning_rate": 4.4534855694856e-05, "loss": 0.041, "step": 89300 }, { "epoch": 1.7847008512849207, "grad_norm": 0.10300150513648987, "learning_rate": 4.452458354267535e-05, "loss": 0.031, "step": 89310 }, { "epoch": 1.7849006834259222, "grad_norm": 0.13665448129177094, "learning_rate": 4.4514311624394455e-05, "loss": 0.0255, "step": 89320 }, { "epoch": 1.7851005155669237, "grad_norm": 0.1079072579741478, "learning_rate": 4.4504039940452075e-05, "loss": 0.0305, "step": 89330 }, { "epoch": 1.7853003477079255, "grad_norm": 0.11828470230102539, "learning_rate": 4.449376849128703e-05, "loss": 0.0261, "step": 89340 }, { "epoch": 1.785500179848927, "grad_norm": 0.1907937079668045, "learning_rate": 4.448349727733805e-05, "loss": 0.0333, "step": 89350 }, { "epoch": 1.7857000119899284, "grad_norm": 0.24203042685985565, "learning_rate": 4.447322629904395e-05, "loss": 0.0286, "step": 89360 }, { "epoch": 1.7858998441309302, "grad_norm": 0.12326392531394958, "learning_rate": 4.446295555684347e-05, "loss": 0.027, "step": 89370 }, { "epoch": 1.7860996762719314, "grad_norm": 0.12964366376399994, "learning_rate": 4.445268505117535e-05, "loss": 0.0325, "step": 89380 }, { "epoch": 1.7862995084129332, "grad_norm": 0.1537885069847107, "learning_rate": 4.444241478247835e-05, "loss": 0.0305, "step": 89390 }, { "epoch": 1.7864993405539347, "grad_norm": 0.15365846455097198, "learning_rate": 4.4432144751191154e-05, "loss": 0.0375, "step": 89400 }, { "epoch": 1.7866991726949362, "grad_norm": 0.11515507847070694, "learning_rate": 4.442187495775252e-05, "loss": 0.0237, "step": 89410 }, { "epoch": 1.7868990048359379, "grad_norm": 0.11521207541227341, "learning_rate": 4.441160540260115e-05, "loss": 0.0263, "step": 89420 }, { "epoch": 1.7870988369769394, "grad_norm": 0.13895121216773987, "learning_rate": 4.440133608617571e-05, "loss": 0.0252, "step": 89430 }, { "epoch": 1.7872986691179409, "grad_norm": 0.10542848706245422, "learning_rate": 4.439106700891491e-05, "loss": 0.0266, "step": 89440 }, { "epoch": 1.7874985012589426, "grad_norm": 0.16967423260211945, "learning_rate": 4.438079817125743e-05, "loss": 0.0336, "step": 89450 }, { "epoch": 1.7876983333999439, "grad_norm": 0.12175252288579941, "learning_rate": 4.437052957364195e-05, "loss": 0.0231, "step": 89460 }, { "epoch": 1.7878981655409456, "grad_norm": 0.1093243807554245, "learning_rate": 4.436026121650708e-05, "loss": 0.0325, "step": 89470 }, { "epoch": 1.7880979976819473, "grad_norm": 0.10999631136655807, "learning_rate": 4.43499931002915e-05, "loss": 0.0353, "step": 89480 }, { "epoch": 1.7882978298229486, "grad_norm": 0.12196862697601318, "learning_rate": 4.433972522543383e-05, "loss": 0.0262, "step": 89490 }, { "epoch": 1.7884976619639503, "grad_norm": 0.1301288604736328, "learning_rate": 4.4329457592372694e-05, "loss": 0.0321, "step": 89500 }, { "epoch": 1.7886974941049518, "grad_norm": 0.12788870930671692, "learning_rate": 4.4319190201546726e-05, "loss": 0.0241, "step": 89510 }, { "epoch": 1.7888973262459533, "grad_norm": 0.14521798491477966, "learning_rate": 4.430892305339451e-05, "loss": 0.0305, "step": 89520 }, { "epoch": 1.789097158386955, "grad_norm": 0.14598503708839417, "learning_rate": 4.429865614835464e-05, "loss": 0.0241, "step": 89530 }, { "epoch": 1.7892969905279565, "grad_norm": 0.16169418394565582, "learning_rate": 4.428838948686572e-05, "loss": 0.0348, "step": 89540 }, { "epoch": 1.789496822668958, "grad_norm": 0.12270349264144897, "learning_rate": 4.42781230693663e-05, "loss": 0.0366, "step": 89550 }, { "epoch": 1.7896966548099598, "grad_norm": 0.1257459968328476, "learning_rate": 4.4267856896294964e-05, "loss": 0.0268, "step": 89560 }, { "epoch": 1.789896486950961, "grad_norm": 0.10066346079111099, "learning_rate": 4.425759096809023e-05, "loss": 0.0243, "step": 89570 }, { "epoch": 1.7900963190919628, "grad_norm": 0.12139308452606201, "learning_rate": 4.424732528519067e-05, "loss": 0.0339, "step": 89580 }, { "epoch": 1.7902961512329643, "grad_norm": 0.1348200887441635, "learning_rate": 4.423705984803481e-05, "loss": 0.0318, "step": 89590 }, { "epoch": 1.7904959833739658, "grad_norm": 0.13096174597740173, "learning_rate": 4.422679465706116e-05, "loss": 0.0277, "step": 89600 }, { "epoch": 1.7906958155149675, "grad_norm": 0.15676654875278473, "learning_rate": 4.421652971270825e-05, "loss": 0.0292, "step": 89610 }, { "epoch": 1.790895647655969, "grad_norm": 0.14980295300483704, "learning_rate": 4.4206265015414544e-05, "loss": 0.029, "step": 89620 }, { "epoch": 1.7910954797969705, "grad_norm": 0.12331878393888474, "learning_rate": 4.4196000565618565e-05, "loss": 0.0333, "step": 89630 }, { "epoch": 1.7912953119379722, "grad_norm": 0.11925332248210907, "learning_rate": 4.4185736363758785e-05, "loss": 0.0329, "step": 89640 }, { "epoch": 1.7914951440789737, "grad_norm": 0.1424787938594818, "learning_rate": 4.417547241027366e-05, "loss": 0.0385, "step": 89650 }, { "epoch": 1.7916949762199752, "grad_norm": 0.19461581110954285, "learning_rate": 4.4165208705601664e-05, "loss": 0.0344, "step": 89660 }, { "epoch": 1.791894808360977, "grad_norm": 0.15083448588848114, "learning_rate": 4.4154945250181206e-05, "loss": 0.0303, "step": 89670 }, { "epoch": 1.7920946405019782, "grad_norm": 0.17162008583545685, "learning_rate": 4.4144682044450766e-05, "loss": 0.03, "step": 89680 }, { "epoch": 1.79229447264298, "grad_norm": 0.12997348606586456, "learning_rate": 4.413441908884877e-05, "loss": 0.0296, "step": 89690 }, { "epoch": 1.7924943047839814, "grad_norm": 0.1538175493478775, "learning_rate": 4.412415638381359e-05, "loss": 0.0297, "step": 89700 }, { "epoch": 1.792694136924983, "grad_norm": 0.11855140328407288, "learning_rate": 4.4113893929783654e-05, "loss": 0.0266, "step": 89710 }, { "epoch": 1.7928939690659846, "grad_norm": 0.13550494611263275, "learning_rate": 4.410363172719737e-05, "loss": 0.0241, "step": 89720 }, { "epoch": 1.7930938012069861, "grad_norm": 0.13070757687091827, "learning_rate": 4.4093369776493096e-05, "loss": 0.0341, "step": 89730 }, { "epoch": 1.7932936333479876, "grad_norm": 0.10778041929006577, "learning_rate": 4.408310807810924e-05, "loss": 0.0335, "step": 89740 }, { "epoch": 1.7934934654889894, "grad_norm": 0.15086032450199127, "learning_rate": 4.407284663248411e-05, "loss": 0.0291, "step": 89750 }, { "epoch": 1.7936932976299909, "grad_norm": 0.1111137792468071, "learning_rate": 4.406258544005608e-05, "loss": 0.0284, "step": 89760 }, { "epoch": 1.7938931297709924, "grad_norm": 0.132050022482872, "learning_rate": 4.405232450126352e-05, "loss": 0.0366, "step": 89770 }, { "epoch": 1.794092961911994, "grad_norm": 0.14665360748767853, "learning_rate": 4.40420638165447e-05, "loss": 0.0252, "step": 89780 }, { "epoch": 1.7942927940529954, "grad_norm": 0.09161935746669769, "learning_rate": 4.4031803386338e-05, "loss": 0.0346, "step": 89790 }, { "epoch": 1.794492626193997, "grad_norm": 0.09486457705497742, "learning_rate": 4.402154321108165e-05, "loss": 0.0257, "step": 89800 }, { "epoch": 1.7946924583349986, "grad_norm": 0.09843848645687103, "learning_rate": 4.401128329121404e-05, "loss": 0.0265, "step": 89810 }, { "epoch": 1.794892290476, "grad_norm": 0.08185133337974548, "learning_rate": 4.4001023627173386e-05, "loss": 0.0317, "step": 89820 }, { "epoch": 1.7950921226170018, "grad_norm": 0.13952401280403137, "learning_rate": 4.399076421939799e-05, "loss": 0.0269, "step": 89830 }, { "epoch": 1.7952919547580033, "grad_norm": 0.10811050236225128, "learning_rate": 4.398050506832608e-05, "loss": 0.0354, "step": 89840 }, { "epoch": 1.7954917868990048, "grad_norm": 0.10199077427387238, "learning_rate": 4.397024617439595e-05, "loss": 0.0294, "step": 89850 }, { "epoch": 1.7956916190400065, "grad_norm": 0.14612168073654175, "learning_rate": 4.395998753804584e-05, "loss": 0.0209, "step": 89860 }, { "epoch": 1.7958914511810078, "grad_norm": 0.11309455335140228, "learning_rate": 4.394972915971396e-05, "loss": 0.0266, "step": 89870 }, { "epoch": 1.7960912833220095, "grad_norm": 0.10487614572048187, "learning_rate": 4.3939471039838544e-05, "loss": 0.0281, "step": 89880 }, { "epoch": 1.7962911154630112, "grad_norm": 0.09780823439359665, "learning_rate": 4.3929213178857764e-05, "loss": 0.0265, "step": 89890 }, { "epoch": 1.7964909476040125, "grad_norm": 0.12863074243068695, "learning_rate": 4.3918955577209864e-05, "loss": 0.0267, "step": 89900 }, { "epoch": 1.7966907797450142, "grad_norm": 0.08393386751413345, "learning_rate": 4.390869823533301e-05, "loss": 0.0252, "step": 89910 }, { "epoch": 1.7968906118860157, "grad_norm": 0.09538228064775467, "learning_rate": 4.3898441153665375e-05, "loss": 0.0248, "step": 89920 }, { "epoch": 1.7970904440270172, "grad_norm": 0.1039099469780922, "learning_rate": 4.388818433264511e-05, "loss": 0.0244, "step": 89930 }, { "epoch": 1.797290276168019, "grad_norm": 0.168653205037117, "learning_rate": 4.387792777271042e-05, "loss": 0.0244, "step": 89940 }, { "epoch": 1.7974901083090205, "grad_norm": 0.13132412731647491, "learning_rate": 4.386767147429938e-05, "loss": 0.0207, "step": 89950 }, { "epoch": 1.797689940450022, "grad_norm": 0.1424172818660736, "learning_rate": 4.385741543785016e-05, "loss": 0.029, "step": 89960 }, { "epoch": 1.7978897725910237, "grad_norm": 0.15368704497814178, "learning_rate": 4.3847159663800866e-05, "loss": 0.0308, "step": 89970 }, { "epoch": 1.798089604732025, "grad_norm": 0.12088573724031448, "learning_rate": 4.3836904152589595e-05, "loss": 0.0257, "step": 89980 }, { "epoch": 1.7982894368730267, "grad_norm": 0.10395285487174988, "learning_rate": 4.3826648904654474e-05, "loss": 0.0323, "step": 89990 }, { "epoch": 1.7984892690140282, "grad_norm": 0.12823523581027985, "learning_rate": 4.381639392043356e-05, "loss": 0.0235, "step": 90000 }, { "epoch": 1.7986891011550297, "grad_norm": 0.12576085329055786, "learning_rate": 4.380613920036496e-05, "loss": 0.029, "step": 90010 }, { "epoch": 1.7988889332960314, "grad_norm": 0.13495321571826935, "learning_rate": 4.3795884744886684e-05, "loss": 0.0229, "step": 90020 }, { "epoch": 1.7990887654370329, "grad_norm": 0.11972242593765259, "learning_rate": 4.3785630554436825e-05, "loss": 0.0286, "step": 90030 }, { "epoch": 1.7992885975780344, "grad_norm": 0.16848929226398468, "learning_rate": 4.377537662945342e-05, "loss": 0.0324, "step": 90040 }, { "epoch": 1.799488429719036, "grad_norm": 0.12024272233247757, "learning_rate": 4.376512297037449e-05, "loss": 0.0335, "step": 90050 }, { "epoch": 1.7996882618600376, "grad_norm": 0.09720485657453537, "learning_rate": 4.375486957763806e-05, "loss": 0.0276, "step": 90060 }, { "epoch": 1.799888094001039, "grad_norm": 0.14390335977077484, "learning_rate": 4.37446164516821e-05, "loss": 0.0335, "step": 90070 }, { "epoch": 1.8000879261420408, "grad_norm": 0.1720091849565506, "learning_rate": 4.373436359294466e-05, "loss": 0.0309, "step": 90080 }, { "epoch": 1.800287758283042, "grad_norm": 0.1204913854598999, "learning_rate": 4.3724111001863697e-05, "loss": 0.0246, "step": 90090 }, { "epoch": 1.8004875904240438, "grad_norm": 0.12045548856258392, "learning_rate": 4.371385867887717e-05, "loss": 0.0247, "step": 90100 }, { "epoch": 1.8006874225650453, "grad_norm": 0.12585502862930298, "learning_rate": 4.370360662442304e-05, "loss": 0.0239, "step": 90110 }, { "epoch": 1.8008872547060468, "grad_norm": 0.11157885938882828, "learning_rate": 4.3693354838939295e-05, "loss": 0.0248, "step": 90120 }, { "epoch": 1.8010870868470485, "grad_norm": 0.1244187280535698, "learning_rate": 4.368310332286383e-05, "loss": 0.0315, "step": 90130 }, { "epoch": 1.80128691898805, "grad_norm": 0.12452187389135361, "learning_rate": 4.367285207663461e-05, "loss": 0.0337, "step": 90140 }, { "epoch": 1.8014867511290515, "grad_norm": 0.0935896709561348, "learning_rate": 4.3662601100689504e-05, "loss": 0.0202, "step": 90150 }, { "epoch": 1.8016865832700533, "grad_norm": 0.12192956358194351, "learning_rate": 4.365235039546642e-05, "loss": 0.0333, "step": 90160 }, { "epoch": 1.8018864154110548, "grad_norm": 0.12042277306318283, "learning_rate": 4.3642099961403304e-05, "loss": 0.0271, "step": 90170 }, { "epoch": 1.8020862475520563, "grad_norm": 0.15576888620853424, "learning_rate": 4.363184979893798e-05, "loss": 0.0364, "step": 90180 }, { "epoch": 1.802286079693058, "grad_norm": 0.10363242775201797, "learning_rate": 4.362159990850834e-05, "loss": 0.0289, "step": 90190 }, { "epoch": 1.8024859118340593, "grad_norm": 0.13879495859146118, "learning_rate": 4.361135029055221e-05, "loss": 0.0301, "step": 90200 }, { "epoch": 1.802685743975061, "grad_norm": 0.1416623890399933, "learning_rate": 4.3601100945507485e-05, "loss": 0.0275, "step": 90210 }, { "epoch": 1.8028855761160625, "grad_norm": 0.1290874183177948, "learning_rate": 4.3590851873811964e-05, "loss": 0.0279, "step": 90220 }, { "epoch": 1.803085408257064, "grad_norm": 0.12951457500457764, "learning_rate": 4.3580603075903485e-05, "loss": 0.0257, "step": 90230 }, { "epoch": 1.8032852403980657, "grad_norm": 0.11113186925649643, "learning_rate": 4.357035455221984e-05, "loss": 0.0259, "step": 90240 }, { "epoch": 1.8034850725390672, "grad_norm": 0.10533063113689423, "learning_rate": 4.356010630319883e-05, "loss": 0.0297, "step": 90250 }, { "epoch": 1.8036849046800687, "grad_norm": 0.1018705666065216, "learning_rate": 4.354985832927827e-05, "loss": 0.0232, "step": 90260 }, { "epoch": 1.8038847368210704, "grad_norm": 0.11769093573093414, "learning_rate": 4.35396106308959e-05, "loss": 0.0286, "step": 90270 }, { "epoch": 1.8040845689620717, "grad_norm": 0.10518038272857666, "learning_rate": 4.3529363208489505e-05, "loss": 0.0278, "step": 90280 }, { "epoch": 1.8042844011030734, "grad_norm": 0.1325443685054779, "learning_rate": 4.351911606249681e-05, "loss": 0.029, "step": 90290 }, { "epoch": 1.8044842332440751, "grad_norm": 0.12928472459316254, "learning_rate": 4.350886919335558e-05, "loss": 0.0337, "step": 90300 }, { "epoch": 1.8046840653850764, "grad_norm": 0.09302125871181488, "learning_rate": 4.349862260150354e-05, "loss": 0.0277, "step": 90310 }, { "epoch": 1.8048838975260781, "grad_norm": 0.1208251342177391, "learning_rate": 4.3488376287378394e-05, "loss": 0.028, "step": 90320 }, { "epoch": 1.8050837296670796, "grad_norm": 0.15598635375499725, "learning_rate": 4.347813025141785e-05, "loss": 0.0278, "step": 90330 }, { "epoch": 1.8052835618080811, "grad_norm": 0.14394810795783997, "learning_rate": 4.3467884494059594e-05, "loss": 0.029, "step": 90340 }, { "epoch": 1.8054833939490829, "grad_norm": 0.11132851988077164, "learning_rate": 4.345763901574132e-05, "loss": 0.0299, "step": 90350 }, { "epoch": 1.8056832260900844, "grad_norm": 0.10265009850263596, "learning_rate": 4.344739381690069e-05, "loss": 0.0273, "step": 90360 }, { "epoch": 1.8058830582310859, "grad_norm": 0.09443282335996628, "learning_rate": 4.3437148897975354e-05, "loss": 0.0265, "step": 90370 }, { "epoch": 1.8060828903720876, "grad_norm": 0.10440194606781006, "learning_rate": 4.342690425940295e-05, "loss": 0.0266, "step": 90380 }, { "epoch": 1.8062827225130889, "grad_norm": 0.12100912630558014, "learning_rate": 4.3416659901621146e-05, "loss": 0.0288, "step": 90390 }, { "epoch": 1.8064825546540906, "grad_norm": 0.13963143527507782, "learning_rate": 4.340641582506752e-05, "loss": 0.0266, "step": 90400 }, { "epoch": 1.806682386795092, "grad_norm": 0.19014447927474976, "learning_rate": 4.339617203017971e-05, "loss": 0.0343, "step": 90410 }, { "epoch": 1.8068822189360936, "grad_norm": 0.12852995097637177, "learning_rate": 4.3385928517395294e-05, "loss": 0.0303, "step": 90420 }, { "epoch": 1.8070820510770953, "grad_norm": 0.14377178251743317, "learning_rate": 4.337568528715184e-05, "loss": 0.0266, "step": 90430 }, { "epoch": 1.8072818832180968, "grad_norm": 0.12103883177042007, "learning_rate": 4.336544233988697e-05, "loss": 0.0293, "step": 90440 }, { "epoch": 1.8074817153590983, "grad_norm": 0.1322290599346161, "learning_rate": 4.33551996760382e-05, "loss": 0.0361, "step": 90450 }, { "epoch": 1.8076815475001, "grad_norm": 0.12062393128871918, "learning_rate": 4.334495729604311e-05, "loss": 0.0234, "step": 90460 }, { "epoch": 1.8078813796411015, "grad_norm": 0.09841908514499664, "learning_rate": 4.3334715200339194e-05, "loss": 0.0294, "step": 90470 }, { "epoch": 1.808081211782103, "grad_norm": 0.1271667182445526, "learning_rate": 4.3324473389364005e-05, "loss": 0.0341, "step": 90480 }, { "epoch": 1.8082810439231047, "grad_norm": 0.11792687326669693, "learning_rate": 4.3314231863555075e-05, "loss": 0.0201, "step": 90490 }, { "epoch": 1.808480876064106, "grad_norm": 0.14145298302173615, "learning_rate": 4.330399062334985e-05, "loss": 0.0413, "step": 90500 }, { "epoch": 1.8086807082051077, "grad_norm": 0.10899323970079422, "learning_rate": 4.329374966918587e-05, "loss": 0.0312, "step": 90510 }, { "epoch": 1.8088805403461092, "grad_norm": 0.10757188498973846, "learning_rate": 4.3283509001500554e-05, "loss": 0.0261, "step": 90520 }, { "epoch": 1.8090803724871107, "grad_norm": 0.13587787747383118, "learning_rate": 4.32732686207314e-05, "loss": 0.0321, "step": 90530 }, { "epoch": 1.8092802046281125, "grad_norm": 0.12408845126628876, "learning_rate": 4.3263028527315874e-05, "loss": 0.0286, "step": 90540 }, { "epoch": 1.809480036769114, "grad_norm": 0.12789982557296753, "learning_rate": 4.3252788721691384e-05, "loss": 0.0196, "step": 90550 }, { "epoch": 1.8096798689101155, "grad_norm": 0.10210862010717392, "learning_rate": 4.324254920429535e-05, "loss": 0.0292, "step": 90560 }, { "epoch": 1.8098797010511172, "grad_norm": 0.17689105868339539, "learning_rate": 4.3232309975565224e-05, "loss": 0.0296, "step": 90570 }, { "epoch": 1.8100795331921187, "grad_norm": 0.1528879851102829, "learning_rate": 4.322207103593837e-05, "loss": 0.0345, "step": 90580 }, { "epoch": 1.8102793653331202, "grad_norm": 0.10951701551675797, "learning_rate": 4.32118323858522e-05, "loss": 0.0302, "step": 90590 }, { "epoch": 1.810479197474122, "grad_norm": 0.11602780222892761, "learning_rate": 4.320159402574406e-05, "loss": 0.0269, "step": 90600 }, { "epoch": 1.8106790296151232, "grad_norm": 0.08410479873418808, "learning_rate": 4.319135595605134e-05, "loss": 0.0426, "step": 90610 }, { "epoch": 1.810878861756125, "grad_norm": 0.1197432354092598, "learning_rate": 4.318111817721139e-05, "loss": 0.0246, "step": 90620 }, { "epoch": 1.8110786938971264, "grad_norm": 0.15438765287399292, "learning_rate": 4.317088068966155e-05, "loss": 0.027, "step": 90630 }, { "epoch": 1.811278526038128, "grad_norm": 0.1663666069507599, "learning_rate": 4.316064349383914e-05, "loss": 0.0261, "step": 90640 }, { "epoch": 1.8114783581791296, "grad_norm": 0.20416441559791565, "learning_rate": 4.315040659018146e-05, "loss": 0.0309, "step": 90650 }, { "epoch": 1.811678190320131, "grad_norm": 0.12066889554262161, "learning_rate": 4.314016997912585e-05, "loss": 0.0319, "step": 90660 }, { "epoch": 1.8118780224611326, "grad_norm": 0.1181730106472969, "learning_rate": 4.3129933661109566e-05, "loss": 0.0333, "step": 90670 }, { "epoch": 1.8120778546021343, "grad_norm": 0.10890799015760422, "learning_rate": 4.3119697636569903e-05, "loss": 0.0218, "step": 90680 }, { "epoch": 1.8122776867431356, "grad_norm": 0.12829115986824036, "learning_rate": 4.3109461905944114e-05, "loss": 0.0231, "step": 90690 }, { "epoch": 1.8124775188841373, "grad_norm": 0.13078901171684265, "learning_rate": 4.309922646966944e-05, "loss": 0.0256, "step": 90700 }, { "epoch": 1.812677351025139, "grad_norm": 0.08444833755493164, "learning_rate": 4.308899132818315e-05, "loss": 0.0393, "step": 90710 }, { "epoch": 1.8128771831661403, "grad_norm": 0.10575835406780243, "learning_rate": 4.307875648192245e-05, "loss": 0.0272, "step": 90720 }, { "epoch": 1.813077015307142, "grad_norm": 0.1269119828939438, "learning_rate": 4.306852193132457e-05, "loss": 0.027, "step": 90730 }, { "epoch": 1.8132768474481435, "grad_norm": 0.1284259408712387, "learning_rate": 4.305828767682667e-05, "loss": 0.0343, "step": 90740 }, { "epoch": 1.813476679589145, "grad_norm": 0.1264958530664444, "learning_rate": 4.304805371886599e-05, "loss": 0.0327, "step": 90750 }, { "epoch": 1.8136765117301468, "grad_norm": 0.11042247712612152, "learning_rate": 4.3037820057879684e-05, "loss": 0.0211, "step": 90760 }, { "epoch": 1.8138763438711483, "grad_norm": 0.1837400197982788, "learning_rate": 4.3027586694304914e-05, "loss": 0.03, "step": 90770 }, { "epoch": 1.8140761760121498, "grad_norm": 0.13450798392295837, "learning_rate": 4.301735362857883e-05, "loss": 0.0306, "step": 90780 }, { "epoch": 1.8142760081531515, "grad_norm": 0.1456632763147354, "learning_rate": 4.300712086113856e-05, "loss": 0.0307, "step": 90790 }, { "epoch": 1.8144758402941528, "grad_norm": 0.160081684589386, "learning_rate": 4.299688839242125e-05, "loss": 0.0274, "step": 90800 }, { "epoch": 1.8146756724351545, "grad_norm": 0.11166314780712128, "learning_rate": 4.298665622286401e-05, "loss": 0.0281, "step": 90810 }, { "epoch": 1.814875504576156, "grad_norm": 0.128315269947052, "learning_rate": 4.297642435290393e-05, "loss": 0.0343, "step": 90820 }, { "epoch": 1.8150753367171575, "grad_norm": 0.1194678544998169, "learning_rate": 4.296619278297807e-05, "loss": 0.0306, "step": 90830 }, { "epoch": 1.8152751688581592, "grad_norm": 0.13214834034442902, "learning_rate": 4.295596151352358e-05, "loss": 0.0342, "step": 90840 }, { "epoch": 1.8154750009991607, "grad_norm": 0.16715426743030548, "learning_rate": 4.294573054497744e-05, "loss": 0.0353, "step": 90850 }, { "epoch": 1.8156748331401622, "grad_norm": 0.18562795221805573, "learning_rate": 4.293549987777675e-05, "loss": 0.0254, "step": 90860 }, { "epoch": 1.815874665281164, "grad_norm": 0.10693120211362839, "learning_rate": 4.292526951235852e-05, "loss": 0.0288, "step": 90870 }, { "epoch": 1.8160744974221654, "grad_norm": 0.11074452847242355, "learning_rate": 4.291503944915977e-05, "loss": 0.0312, "step": 90880 }, { "epoch": 1.816274329563167, "grad_norm": 0.12726470828056335, "learning_rate": 4.290480968861754e-05, "loss": 0.0298, "step": 90890 }, { "epoch": 1.8164741617041686, "grad_norm": 0.12097801268100739, "learning_rate": 4.289458023116879e-05, "loss": 0.0269, "step": 90900 }, { "epoch": 1.81667399384517, "grad_norm": 0.118911974132061, "learning_rate": 4.288435107725054e-05, "loss": 0.0243, "step": 90910 }, { "epoch": 1.8168738259861716, "grad_norm": 0.11415277421474457, "learning_rate": 4.287412222729971e-05, "loss": 0.0236, "step": 90920 }, { "epoch": 1.8170736581271731, "grad_norm": 0.1459643691778183, "learning_rate": 4.2863893681753294e-05, "loss": 0.0251, "step": 90930 }, { "epoch": 1.8172734902681746, "grad_norm": 0.12146797776222229, "learning_rate": 4.285366544104825e-05, "loss": 0.0284, "step": 90940 }, { "epoch": 1.8174733224091764, "grad_norm": 0.11693032830953598, "learning_rate": 4.284343750562148e-05, "loss": 0.0344, "step": 90950 }, { "epoch": 1.8176731545501779, "grad_norm": 0.14088141918182373, "learning_rate": 4.2833209875909926e-05, "loss": 0.0297, "step": 90960 }, { "epoch": 1.8178729866911794, "grad_norm": 0.1521829217672348, "learning_rate": 4.2822982552350446e-05, "loss": 0.0193, "step": 90970 }, { "epoch": 1.818072818832181, "grad_norm": 0.09967118501663208, "learning_rate": 4.281275553537999e-05, "loss": 0.036, "step": 90980 }, { "epoch": 1.8182726509731826, "grad_norm": 0.11454807966947556, "learning_rate": 4.2802528825435415e-05, "loss": 0.027, "step": 90990 }, { "epoch": 1.818472483114184, "grad_norm": 0.15917205810546875, "learning_rate": 4.279230242295359e-05, "loss": 0.0309, "step": 91000 }, { "epoch": 1.8186723152551858, "grad_norm": 0.12283647060394287, "learning_rate": 4.278207632837134e-05, "loss": 0.0303, "step": 91010 }, { "epoch": 1.818872147396187, "grad_norm": 0.10498616099357605, "learning_rate": 4.277185054212555e-05, "loss": 0.0349, "step": 91020 }, { "epoch": 1.8190719795371888, "grad_norm": 0.08761221170425415, "learning_rate": 4.2761625064653024e-05, "loss": 0.0297, "step": 91030 }, { "epoch": 1.8192718116781903, "grad_norm": 0.14714518189430237, "learning_rate": 4.2751399896390576e-05, "loss": 0.0252, "step": 91040 }, { "epoch": 1.8194716438191918, "grad_norm": 0.12599702179431915, "learning_rate": 4.2741175037775e-05, "loss": 0.0268, "step": 91050 }, { "epoch": 1.8196714759601935, "grad_norm": 0.10495202988386154, "learning_rate": 4.273095048924311e-05, "loss": 0.0305, "step": 91060 }, { "epoch": 1.819871308101195, "grad_norm": 0.13766150176525116, "learning_rate": 4.272072625123165e-05, "loss": 0.0236, "step": 91070 }, { "epoch": 1.8200711402421965, "grad_norm": 0.12935036420822144, "learning_rate": 4.27105023241774e-05, "loss": 0.0333, "step": 91080 }, { "epoch": 1.8202709723831982, "grad_norm": 0.1128113642334938, "learning_rate": 4.2700278708517095e-05, "loss": 0.0259, "step": 91090 }, { "epoch": 1.8204708045241995, "grad_norm": 0.10278142988681793, "learning_rate": 4.2690055404687456e-05, "loss": 0.0291, "step": 91100 }, { "epoch": 1.8206706366652012, "grad_norm": 0.19634713232517242, "learning_rate": 4.267983241312525e-05, "loss": 0.0297, "step": 91110 }, { "epoch": 1.8208704688062027, "grad_norm": 0.12729512155056, "learning_rate": 4.266960973426714e-05, "loss": 0.0288, "step": 91120 }, { "epoch": 1.8210703009472042, "grad_norm": 0.12948419153690338, "learning_rate": 4.265938736854985e-05, "loss": 0.0288, "step": 91130 }, { "epoch": 1.821270133088206, "grad_norm": 0.12819893658161163, "learning_rate": 4.264916531641002e-05, "loss": 0.0264, "step": 91140 }, { "epoch": 1.8214699652292075, "grad_norm": 0.10294382274150848, "learning_rate": 4.2638943578284355e-05, "loss": 0.03, "step": 91150 }, { "epoch": 1.821669797370209, "grad_norm": 0.13569264113903046, "learning_rate": 4.2628722154609514e-05, "loss": 0.0293, "step": 91160 }, { "epoch": 1.8218696295112107, "grad_norm": 0.10886429995298386, "learning_rate": 4.2618501045822105e-05, "loss": 0.029, "step": 91170 }, { "epoch": 1.8220694616522122, "grad_norm": 0.10611962527036667, "learning_rate": 4.260828025235879e-05, "loss": 0.0178, "step": 91180 }, { "epoch": 1.8222692937932137, "grad_norm": 0.10612271726131439, "learning_rate": 4.259805977465613e-05, "loss": 0.0259, "step": 91190 }, { "epoch": 1.8224691259342154, "grad_norm": 0.10438892990350723, "learning_rate": 4.2587839613150785e-05, "loss": 0.0227, "step": 91200 }, { "epoch": 1.8226689580752167, "grad_norm": 0.1337542086839676, "learning_rate": 4.257761976827931e-05, "loss": 0.0274, "step": 91210 }, { "epoch": 1.8228687902162184, "grad_norm": 0.0887032002210617, "learning_rate": 4.256740024047828e-05, "loss": 0.0299, "step": 91220 }, { "epoch": 1.82306862235722, "grad_norm": 0.11176162958145142, "learning_rate": 4.2557181030184244e-05, "loss": 0.0283, "step": 91230 }, { "epoch": 1.8232684544982214, "grad_norm": 0.10699265450239182, "learning_rate": 4.254696213783378e-05, "loss": 0.0291, "step": 91240 }, { "epoch": 1.8234682866392231, "grad_norm": 0.15350769460201263, "learning_rate": 4.25367435638634e-05, "loss": 0.0319, "step": 91250 }, { "epoch": 1.8236681187802246, "grad_norm": 0.14010581374168396, "learning_rate": 4.252652530870963e-05, "loss": 0.0265, "step": 91260 }, { "epoch": 1.823867950921226, "grad_norm": 0.1371365189552307, "learning_rate": 4.251630737280896e-05, "loss": 0.0187, "step": 91270 }, { "epoch": 1.8240677830622278, "grad_norm": 0.20980054140090942, "learning_rate": 4.250608975659788e-05, "loss": 0.0218, "step": 91280 }, { "epoch": 1.8242676152032293, "grad_norm": 0.13486336171627045, "learning_rate": 4.2495872460512905e-05, "loss": 0.0197, "step": 91290 }, { "epoch": 1.8244674473442308, "grad_norm": 0.10765492916107178, "learning_rate": 4.2485655484990455e-05, "loss": 0.0199, "step": 91300 }, { "epoch": 1.8246672794852326, "grad_norm": 0.11765795201063156, "learning_rate": 4.247543883046701e-05, "loss": 0.0308, "step": 91310 }, { "epoch": 1.8248671116262338, "grad_norm": 0.09253950417041779, "learning_rate": 4.2465222497378975e-05, "loss": 0.0242, "step": 91320 }, { "epoch": 1.8250669437672355, "grad_norm": 0.1305391490459442, "learning_rate": 4.245500648616281e-05, "loss": 0.0265, "step": 91330 }, { "epoch": 1.825266775908237, "grad_norm": 0.12309679388999939, "learning_rate": 4.2444790797254905e-05, "loss": 0.0298, "step": 91340 }, { "epoch": 1.8254666080492385, "grad_norm": 0.12397949397563934, "learning_rate": 4.243457543109166e-05, "loss": 0.0275, "step": 91350 }, { "epoch": 1.8256664401902403, "grad_norm": 0.13612373173236847, "learning_rate": 4.242436038810946e-05, "loss": 0.0224, "step": 91360 }, { "epoch": 1.8258662723312418, "grad_norm": 0.15182055532932281, "learning_rate": 4.241414566874464e-05, "loss": 0.0291, "step": 91370 }, { "epoch": 1.8260661044722433, "grad_norm": 0.09262362122535706, "learning_rate": 4.24039312734336e-05, "loss": 0.0254, "step": 91380 }, { "epoch": 1.826265936613245, "grad_norm": 0.11810516566038132, "learning_rate": 4.239371720261267e-05, "loss": 0.0232, "step": 91390 }, { "epoch": 1.8264657687542465, "grad_norm": 0.11695057153701782, "learning_rate": 4.238350345671815e-05, "loss": 0.0311, "step": 91400 }, { "epoch": 1.826665600895248, "grad_norm": 0.10540001839399338, "learning_rate": 4.2373290036186375e-05, "loss": 0.0332, "step": 91410 }, { "epoch": 1.8268654330362497, "grad_norm": 0.11416575312614441, "learning_rate": 4.2363076941453636e-05, "loss": 0.0288, "step": 91420 }, { "epoch": 1.827065265177251, "grad_norm": 0.11234075576066971, "learning_rate": 4.235286417295624e-05, "loss": 0.0242, "step": 91430 }, { "epoch": 1.8272650973182527, "grad_norm": 0.10983528941869736, "learning_rate": 4.2342651731130423e-05, "loss": 0.0243, "step": 91440 }, { "epoch": 1.8274649294592542, "grad_norm": 0.11253059655427933, "learning_rate": 4.2332439616412474e-05, "loss": 0.0279, "step": 91450 }, { "epoch": 1.8276647616002557, "grad_norm": 0.12652519345283508, "learning_rate": 4.2322227829238594e-05, "loss": 0.0316, "step": 91460 }, { "epoch": 1.8278645937412574, "grad_norm": 0.11915018409490585, "learning_rate": 4.231201637004504e-05, "loss": 0.0313, "step": 91470 }, { "epoch": 1.828064425882259, "grad_norm": 0.11833319813013077, "learning_rate": 4.230180523926805e-05, "loss": 0.0296, "step": 91480 }, { "epoch": 1.8282642580232604, "grad_norm": 0.14179988205432892, "learning_rate": 4.229159443734377e-05, "loss": 0.0385, "step": 91490 }, { "epoch": 1.8284640901642621, "grad_norm": 0.09519054740667343, "learning_rate": 4.228138396470841e-05, "loss": 0.0232, "step": 91500 }, { "epoch": 1.8286639223052634, "grad_norm": 0.12259525805711746, "learning_rate": 4.227117382179817e-05, "loss": 0.0332, "step": 91510 }, { "epoch": 1.8288637544462651, "grad_norm": 0.1292862743139267, "learning_rate": 4.2260964009049174e-05, "loss": 0.0265, "step": 91520 }, { "epoch": 1.8290635865872666, "grad_norm": 0.162774920463562, "learning_rate": 4.2250754526897597e-05, "loss": 0.0266, "step": 91530 }, { "epoch": 1.8292634187282681, "grad_norm": 0.11448495835065842, "learning_rate": 4.2240545375779526e-05, "loss": 0.0293, "step": 91540 }, { "epoch": 1.8294632508692699, "grad_norm": 0.15398502349853516, "learning_rate": 4.22303365561311e-05, "loss": 0.0253, "step": 91550 }, { "epoch": 1.8296630830102714, "grad_norm": 0.11455198377370834, "learning_rate": 4.222012806838844e-05, "loss": 0.027, "step": 91560 }, { "epoch": 1.8298629151512729, "grad_norm": 0.1297534853219986, "learning_rate": 4.220991991298762e-05, "loss": 0.026, "step": 91570 }, { "epoch": 1.8300627472922746, "grad_norm": 0.1258143186569214, "learning_rate": 4.219971209036471e-05, "loss": 0.0271, "step": 91580 }, { "epoch": 1.830262579433276, "grad_norm": 0.11144956946372986, "learning_rate": 4.2189504600955745e-05, "loss": 0.0334, "step": 91590 }, { "epoch": 1.8304624115742776, "grad_norm": 0.11569244414567947, "learning_rate": 4.217929744519681e-05, "loss": 0.0208, "step": 91600 }, { "epoch": 1.8306622437152793, "grad_norm": 0.13479210436344147, "learning_rate": 4.216909062352393e-05, "loss": 0.0317, "step": 91610 }, { "epoch": 1.8308620758562806, "grad_norm": 0.1446758508682251, "learning_rate": 4.215888413637311e-05, "loss": 0.0288, "step": 91620 }, { "epoch": 1.8310619079972823, "grad_norm": 0.12714813649654388, "learning_rate": 4.214867798418036e-05, "loss": 0.0338, "step": 91630 }, { "epoch": 1.8312617401382838, "grad_norm": 0.1194242537021637, "learning_rate": 4.213847216738164e-05, "loss": 0.0213, "step": 91640 }, { "epoch": 1.8314615722792853, "grad_norm": 0.13112129271030426, "learning_rate": 4.212826668641295e-05, "loss": 0.0267, "step": 91650 }, { "epoch": 1.831661404420287, "grad_norm": 0.12754979729652405, "learning_rate": 4.211806154171027e-05, "loss": 0.0273, "step": 91660 }, { "epoch": 1.8318612365612885, "grad_norm": 0.19243767857551575, "learning_rate": 4.210785673370951e-05, "loss": 0.0369, "step": 91670 }, { "epoch": 1.83206106870229, "grad_norm": 0.14071917533874512, "learning_rate": 4.20976522628466e-05, "loss": 0.0314, "step": 91680 }, { "epoch": 1.8322609008432917, "grad_norm": 0.2290761023759842, "learning_rate": 4.208744812955748e-05, "loss": 0.0217, "step": 91690 }, { "epoch": 1.8324607329842932, "grad_norm": 0.15149424970149994, "learning_rate": 4.207724433427804e-05, "loss": 0.0366, "step": 91700 }, { "epoch": 1.8326605651252947, "grad_norm": 0.12583579123020172, "learning_rate": 4.206704087744418e-05, "loss": 0.0305, "step": 91710 }, { "epoch": 1.8328603972662965, "grad_norm": 0.11164482682943344, "learning_rate": 4.205683775949174e-05, "loss": 0.028, "step": 91720 }, { "epoch": 1.8330602294072977, "grad_norm": 0.10951835662126541, "learning_rate": 4.2046634980856595e-05, "loss": 0.0312, "step": 91730 }, { "epoch": 1.8332600615482995, "grad_norm": 0.1449253261089325, "learning_rate": 4.203643254197461e-05, "loss": 0.0338, "step": 91740 }, { "epoch": 1.833459893689301, "grad_norm": 0.13392123579978943, "learning_rate": 4.20262304432816e-05, "loss": 0.023, "step": 91750 }, { "epoch": 1.8336597258303025, "grad_norm": 0.13952650129795074, "learning_rate": 4.201602868521337e-05, "loss": 0.0294, "step": 91760 }, { "epoch": 1.8338595579713042, "grad_norm": 0.12759171426296234, "learning_rate": 4.200582726820571e-05, "loss": 0.0246, "step": 91770 }, { "epoch": 1.8340593901123057, "grad_norm": 0.1218726709485054, "learning_rate": 4.199562619269444e-05, "loss": 0.0289, "step": 91780 }, { "epoch": 1.8342592222533072, "grad_norm": 0.11055875569581985, "learning_rate": 4.1985425459115325e-05, "loss": 0.0208, "step": 91790 }, { "epoch": 1.834459054394309, "grad_norm": 0.08464408665895462, "learning_rate": 4.1975225067904096e-05, "loss": 0.0305, "step": 91800 }, { "epoch": 1.8346588865353102, "grad_norm": 0.09882720559835434, "learning_rate": 4.196502501949652e-05, "loss": 0.0306, "step": 91810 }, { "epoch": 1.834858718676312, "grad_norm": 0.11456514149904251, "learning_rate": 4.1954825314328284e-05, "loss": 0.0205, "step": 91820 }, { "epoch": 1.8350585508173136, "grad_norm": 0.14920499920845032, "learning_rate": 4.194462595283516e-05, "loss": 0.0312, "step": 91830 }, { "epoch": 1.835258382958315, "grad_norm": 0.12813468277454376, "learning_rate": 4.193442693545282e-05, "loss": 0.0207, "step": 91840 }, { "epoch": 1.8354582150993166, "grad_norm": 0.14687125384807587, "learning_rate": 4.1924228262616945e-05, "loss": 0.0252, "step": 91850 }, { "epoch": 1.8356580472403181, "grad_norm": 0.14492757618427277, "learning_rate": 4.191402993476318e-05, "loss": 0.0243, "step": 91860 }, { "epoch": 1.8358578793813196, "grad_norm": 0.13912823796272278, "learning_rate": 4.190383195232721e-05, "loss": 0.0291, "step": 91870 }, { "epoch": 1.8360577115223213, "grad_norm": 0.12021230161190033, "learning_rate": 4.1893634315744685e-05, "loss": 0.0283, "step": 91880 }, { "epoch": 1.8362575436633228, "grad_norm": 0.13835196197032928, "learning_rate": 4.188343702545119e-05, "loss": 0.0253, "step": 91890 }, { "epoch": 1.8364573758043243, "grad_norm": 0.11306343972682953, "learning_rate": 4.187324008188238e-05, "loss": 0.0264, "step": 91900 }, { "epoch": 1.836657207945326, "grad_norm": 0.13657744228839874, "learning_rate": 4.186304348547379e-05, "loss": 0.0305, "step": 91910 }, { "epoch": 1.8368570400863273, "grad_norm": 0.11257422715425491, "learning_rate": 4.1852847236661055e-05, "loss": 0.0292, "step": 91920 }, { "epoch": 1.837056872227329, "grad_norm": 0.15696460008621216, "learning_rate": 4.184265133587972e-05, "loss": 0.0312, "step": 91930 }, { "epoch": 1.8372567043683306, "grad_norm": 0.12131179124116898, "learning_rate": 4.1832455783565324e-05, "loss": 0.0349, "step": 91940 }, { "epoch": 1.837456536509332, "grad_norm": 0.11315161734819412, "learning_rate": 4.18222605801534e-05, "loss": 0.0194, "step": 91950 }, { "epoch": 1.8376563686503338, "grad_norm": 0.12891662120819092, "learning_rate": 4.181206572607952e-05, "loss": 0.0335, "step": 91960 }, { "epoch": 1.8378562007913353, "grad_norm": 0.10212162137031555, "learning_rate": 4.180187122177912e-05, "loss": 0.0332, "step": 91970 }, { "epoch": 1.8380560329323368, "grad_norm": 0.14076797664165497, "learning_rate": 4.179167706768774e-05, "loss": 0.0232, "step": 91980 }, { "epoch": 1.8382558650733385, "grad_norm": 0.1248447373509407, "learning_rate": 4.1781483264240825e-05, "loss": 0.0232, "step": 91990 }, { "epoch": 1.83845569721434, "grad_norm": 0.12871122360229492, "learning_rate": 4.177128981187384e-05, "loss": 0.0279, "step": 92000 }, { "epoch": 1.8386555293553415, "grad_norm": 0.1615978479385376, "learning_rate": 4.176109671102225e-05, "loss": 0.0283, "step": 92010 }, { "epoch": 1.8388553614963432, "grad_norm": 0.0836961641907692, "learning_rate": 4.175090396212147e-05, "loss": 0.025, "step": 92020 }, { "epoch": 1.8390551936373445, "grad_norm": 0.12798497080802917, "learning_rate": 4.174071156560693e-05, "loss": 0.0311, "step": 92030 }, { "epoch": 1.8392550257783462, "grad_norm": 0.11289850622415543, "learning_rate": 4.1730519521913996e-05, "loss": 0.0286, "step": 92040 }, { "epoch": 1.8394548579193477, "grad_norm": 0.128285750746727, "learning_rate": 4.172032783147808e-05, "loss": 0.03, "step": 92050 }, { "epoch": 1.8396546900603492, "grad_norm": 0.12340116500854492, "learning_rate": 4.1710136494734574e-05, "loss": 0.0243, "step": 92060 }, { "epoch": 1.839854522201351, "grad_norm": 0.11405684053897858, "learning_rate": 4.169994551211879e-05, "loss": 0.0258, "step": 92070 }, { "epoch": 1.8400543543423524, "grad_norm": 0.10831686109304428, "learning_rate": 4.168975488406609e-05, "loss": 0.0224, "step": 92080 }, { "epoch": 1.840254186483354, "grad_norm": 0.1297626793384552, "learning_rate": 4.167956461101178e-05, "loss": 0.0316, "step": 92090 }, { "epoch": 1.8404540186243556, "grad_norm": 0.1286879926919937, "learning_rate": 4.166937469339119e-05, "loss": 0.0309, "step": 92100 }, { "epoch": 1.8406538507653571, "grad_norm": 0.13183680176734924, "learning_rate": 4.1659185131639635e-05, "loss": 0.035, "step": 92110 }, { "epoch": 1.8408536829063586, "grad_norm": 0.09719675034284592, "learning_rate": 4.1648995926192344e-05, "loss": 0.0222, "step": 92120 }, { "epoch": 1.8410535150473604, "grad_norm": 0.14852623641490936, "learning_rate": 4.16388070774846e-05, "loss": 0.025, "step": 92130 }, { "epoch": 1.8412533471883616, "grad_norm": 0.1266089379787445, "learning_rate": 4.1628618585951676e-05, "loss": 0.0302, "step": 92140 }, { "epoch": 1.8414531793293634, "grad_norm": 0.11933310329914093, "learning_rate": 4.161843045202878e-05, "loss": 0.0324, "step": 92150 }, { "epoch": 1.8416530114703649, "grad_norm": 0.14417023956775665, "learning_rate": 4.1608242676151146e-05, "loss": 0.0322, "step": 92160 }, { "epoch": 1.8418528436113664, "grad_norm": 0.1613510549068451, "learning_rate": 4.159805525875396e-05, "loss": 0.0356, "step": 92170 }, { "epoch": 1.842052675752368, "grad_norm": 0.1108141764998436, "learning_rate": 4.1587868200272403e-05, "loss": 0.031, "step": 92180 }, { "epoch": 1.8422525078933696, "grad_norm": 0.11217659711837769, "learning_rate": 4.157768150114169e-05, "loss": 0.0291, "step": 92190 }, { "epoch": 1.842452340034371, "grad_norm": 0.11219644546508789, "learning_rate": 4.156749516179694e-05, "loss": 0.0328, "step": 92200 }, { "epoch": 1.8426521721753728, "grad_norm": 0.1422831416130066, "learning_rate": 4.155730918267331e-05, "loss": 0.0319, "step": 92210 }, { "epoch": 1.842852004316374, "grad_norm": 0.10707996040582657, "learning_rate": 4.154712356420589e-05, "loss": 0.0267, "step": 92220 }, { "epoch": 1.8430518364573758, "grad_norm": 0.10786256194114685, "learning_rate": 4.153693830682986e-05, "loss": 0.0215, "step": 92230 }, { "epoch": 1.8432516685983775, "grad_norm": 0.11896611005067825, "learning_rate": 4.152675341098027e-05, "loss": 0.0294, "step": 92240 }, { "epoch": 1.8434515007393788, "grad_norm": 0.1092914268374443, "learning_rate": 4.151656887709222e-05, "loss": 0.0224, "step": 92250 }, { "epoch": 1.8436513328803805, "grad_norm": 0.13766026496887207, "learning_rate": 4.1506384705600735e-05, "loss": 0.0279, "step": 92260 }, { "epoch": 1.843851165021382, "grad_norm": 0.10098900645971298, "learning_rate": 4.1496200896940905e-05, "loss": 0.023, "step": 92270 }, { "epoch": 1.8440509971623835, "grad_norm": 0.1542755365371704, "learning_rate": 4.1486017451547764e-05, "loss": 0.0272, "step": 92280 }, { "epoch": 1.8442508293033852, "grad_norm": 0.13749545812606812, "learning_rate": 4.1475834369856306e-05, "loss": 0.02, "step": 92290 }, { "epoch": 1.8444506614443867, "grad_norm": 0.15949133038520813, "learning_rate": 4.1465651652301554e-05, "loss": 0.0309, "step": 92300 }, { "epoch": 1.8446504935853882, "grad_norm": 0.14177441596984863, "learning_rate": 4.1455469299318464e-05, "loss": 0.0307, "step": 92310 }, { "epoch": 1.84485032572639, "grad_norm": 0.20590685307979584, "learning_rate": 4.1445287311342036e-05, "loss": 0.0314, "step": 92320 }, { "epoch": 1.8450501578673912, "grad_norm": 0.13645578920841217, "learning_rate": 4.143510568880723e-05, "loss": 0.0392, "step": 92330 }, { "epoch": 1.845249990008393, "grad_norm": 0.13751216232776642, "learning_rate": 4.142492443214896e-05, "loss": 0.0253, "step": 92340 }, { "epoch": 1.8454498221493945, "grad_norm": 0.113344706594944, "learning_rate": 4.1414743541802164e-05, "loss": 0.0161, "step": 92350 }, { "epoch": 1.845649654290396, "grad_norm": 0.1495102047920227, "learning_rate": 4.140456301820177e-05, "loss": 0.0297, "step": 92360 }, { "epoch": 1.8458494864313977, "grad_norm": 0.12059330940246582, "learning_rate": 4.1394382861782646e-05, "loss": 0.032, "step": 92370 }, { "epoch": 1.8460493185723992, "grad_norm": 0.12616905570030212, "learning_rate": 4.1384203072979685e-05, "loss": 0.0216, "step": 92380 }, { "epoch": 1.8462491507134007, "grad_norm": 0.11833105236291885, "learning_rate": 4.1374023652227726e-05, "loss": 0.0343, "step": 92390 }, { "epoch": 1.8464489828544024, "grad_norm": 0.11856743693351746, "learning_rate": 4.136384459996163e-05, "loss": 0.0359, "step": 92400 }, { "epoch": 1.846648814995404, "grad_norm": 0.1371370255947113, "learning_rate": 4.135366591661624e-05, "loss": 0.0332, "step": 92410 }, { "epoch": 1.8468486471364054, "grad_norm": 0.1641896814107895, "learning_rate": 4.1343487602626347e-05, "loss": 0.0364, "step": 92420 }, { "epoch": 1.8470484792774071, "grad_norm": 0.1281243860721588, "learning_rate": 4.1333309658426775e-05, "loss": 0.0269, "step": 92430 }, { "epoch": 1.8472483114184084, "grad_norm": 0.12275025248527527, "learning_rate": 4.132313208445227e-05, "loss": 0.0271, "step": 92440 }, { "epoch": 1.8474481435594101, "grad_norm": 0.14922739565372467, "learning_rate": 4.131295488113764e-05, "loss": 0.0318, "step": 92450 }, { "epoch": 1.8476479757004116, "grad_norm": 0.11436482518911362, "learning_rate": 4.130277804891762e-05, "loss": 0.0285, "step": 92460 }, { "epoch": 1.8478478078414131, "grad_norm": 0.13994145393371582, "learning_rate": 4.1292601588226944e-05, "loss": 0.0313, "step": 92470 }, { "epoch": 1.8480476399824148, "grad_norm": 0.1452515870332718, "learning_rate": 4.1282425499500335e-05, "loss": 0.0246, "step": 92480 }, { "epoch": 1.8482474721234163, "grad_norm": 0.12039213627576828, "learning_rate": 4.127224978317247e-05, "loss": 0.0263, "step": 92490 }, { "epoch": 1.8484473042644178, "grad_norm": 0.124017633497715, "learning_rate": 4.126207443967808e-05, "loss": 0.0314, "step": 92500 }, { "epoch": 1.8486471364054196, "grad_norm": 0.09835314750671387, "learning_rate": 4.1251899469451824e-05, "loss": 0.0227, "step": 92510 }, { "epoch": 1.848846968546421, "grad_norm": 0.15851838886737823, "learning_rate": 4.124172487292834e-05, "loss": 0.0263, "step": 92520 }, { "epoch": 1.8490468006874226, "grad_norm": 0.10124389082193375, "learning_rate": 4.123155065054227e-05, "loss": 0.027, "step": 92530 }, { "epoch": 1.8492466328284243, "grad_norm": 0.13934439420700073, "learning_rate": 4.122137680272827e-05, "loss": 0.0244, "step": 92540 }, { "epoch": 1.8494464649694256, "grad_norm": 0.12924347817897797, "learning_rate": 4.1211203329920916e-05, "loss": 0.0288, "step": 92550 }, { "epoch": 1.8496462971104273, "grad_norm": 0.09852702170610428, "learning_rate": 4.120103023255482e-05, "loss": 0.0191, "step": 92560 }, { "epoch": 1.8498461292514288, "grad_norm": 0.10289454460144043, "learning_rate": 4.119085751106454e-05, "loss": 0.0354, "step": 92570 }, { "epoch": 1.8500459613924303, "grad_norm": 0.0992170125246048, "learning_rate": 4.118068516588463e-05, "loss": 0.0306, "step": 92580 }, { "epoch": 1.850245793533432, "grad_norm": 0.13391809165477753, "learning_rate": 4.117051319744966e-05, "loss": 0.0303, "step": 92590 }, { "epoch": 1.8504456256744335, "grad_norm": 0.11814083904027939, "learning_rate": 4.116034160619416e-05, "loss": 0.0312, "step": 92600 }, { "epoch": 1.850645457815435, "grad_norm": 0.11145570129156113, "learning_rate": 4.115017039255261e-05, "loss": 0.0396, "step": 92610 }, { "epoch": 1.8508452899564367, "grad_norm": 0.12913034856319427, "learning_rate": 4.113999955695952e-05, "loss": 0.0325, "step": 92620 }, { "epoch": 1.851045122097438, "grad_norm": 0.13347573578357697, "learning_rate": 4.1129829099849396e-05, "loss": 0.0241, "step": 92630 }, { "epoch": 1.8512449542384397, "grad_norm": 0.12385397404432297, "learning_rate": 4.1119659021656654e-05, "loss": 0.0299, "step": 92640 }, { "epoch": 1.8514447863794414, "grad_norm": 0.12247871607542038, "learning_rate": 4.110948932281579e-05, "loss": 0.0278, "step": 92650 }, { "epoch": 1.8516446185204427, "grad_norm": 0.12533850967884064, "learning_rate": 4.109932000376119e-05, "loss": 0.0258, "step": 92660 }, { "epoch": 1.8518444506614444, "grad_norm": 0.11520995199680328, "learning_rate": 4.1089151064927277e-05, "loss": 0.0284, "step": 92670 }, { "epoch": 1.852044282802446, "grad_norm": 0.12431570142507553, "learning_rate": 4.107898250674849e-05, "loss": 0.0291, "step": 92680 }, { "epoch": 1.8522441149434474, "grad_norm": 0.14289972186088562, "learning_rate": 4.106881432965916e-05, "loss": 0.03, "step": 92690 }, { "epoch": 1.8524439470844491, "grad_norm": 0.140253484249115, "learning_rate": 4.105864653409369e-05, "loss": 0.0325, "step": 92700 }, { "epoch": 1.8526437792254506, "grad_norm": 0.10117578506469727, "learning_rate": 4.10484791204864e-05, "loss": 0.0308, "step": 92710 }, { "epoch": 1.8528436113664521, "grad_norm": 0.1328028440475464, "learning_rate": 4.1038312089271644e-05, "loss": 0.0247, "step": 92720 }, { "epoch": 1.8530434435074539, "grad_norm": 0.117359459400177, "learning_rate": 4.1028145440883736e-05, "loss": 0.0301, "step": 92730 }, { "epoch": 1.8532432756484551, "grad_norm": 0.11295473575592041, "learning_rate": 4.1017979175756974e-05, "loss": 0.0267, "step": 92740 }, { "epoch": 1.8534431077894569, "grad_norm": 0.12303134053945541, "learning_rate": 4.100781329432564e-05, "loss": 0.0307, "step": 92750 }, { "epoch": 1.8536429399304584, "grad_norm": 0.11278994381427765, "learning_rate": 4.099764779702399e-05, "loss": 0.0317, "step": 92760 }, { "epoch": 1.8538427720714599, "grad_norm": 0.15040400624275208, "learning_rate": 4.09874826842863e-05, "loss": 0.0277, "step": 92770 }, { "epoch": 1.8540426042124616, "grad_norm": 0.12739285826683044, "learning_rate": 4.0977317956546804e-05, "loss": 0.0294, "step": 92780 }, { "epoch": 1.854242436353463, "grad_norm": 0.14836962521076202, "learning_rate": 4.096715361423969e-05, "loss": 0.0218, "step": 92790 }, { "epoch": 1.8544422684944646, "grad_norm": 0.11230252683162689, "learning_rate": 4.095698965779917e-05, "loss": 0.0262, "step": 92800 }, { "epoch": 1.8546421006354663, "grad_norm": 0.11477471143007278, "learning_rate": 4.094682608765947e-05, "loss": 0.0237, "step": 92810 }, { "epoch": 1.8548419327764678, "grad_norm": 0.12290248274803162, "learning_rate": 4.093666290425471e-05, "loss": 0.0253, "step": 92820 }, { "epoch": 1.8550417649174693, "grad_norm": 0.11784891784191132, "learning_rate": 4.092650010801908e-05, "loss": 0.0215, "step": 92830 }, { "epoch": 1.855241597058471, "grad_norm": 0.13586245477199554, "learning_rate": 4.0916337699386676e-05, "loss": 0.0255, "step": 92840 }, { "epoch": 1.8554414291994723, "grad_norm": 0.15920117497444153, "learning_rate": 4.090617567879163e-05, "loss": 0.0278, "step": 92850 }, { "epoch": 1.855641261340474, "grad_norm": 0.18600791692733765, "learning_rate": 4.089601404666808e-05, "loss": 0.0246, "step": 92860 }, { "epoch": 1.8558410934814755, "grad_norm": 0.1284942924976349, "learning_rate": 4.088585280345007e-05, "loss": 0.0261, "step": 92870 }, { "epoch": 1.856040925622477, "grad_norm": 0.15852944552898407, "learning_rate": 4.08756919495717e-05, "loss": 0.0345, "step": 92880 }, { "epoch": 1.8562407577634787, "grad_norm": 0.13767372071743011, "learning_rate": 4.086553148546698e-05, "loss": 0.022, "step": 92890 }, { "epoch": 1.8564405899044802, "grad_norm": 0.1181284487247467, "learning_rate": 4.0855371411569984e-05, "loss": 0.025, "step": 92900 }, { "epoch": 1.8566404220454817, "grad_norm": 0.11536972224712372, "learning_rate": 4.0845211728314744e-05, "loss": 0.0327, "step": 92910 }, { "epoch": 1.8568402541864835, "grad_norm": 0.14416302740573883, "learning_rate": 4.0835052436135225e-05, "loss": 0.0283, "step": 92920 }, { "epoch": 1.857040086327485, "grad_norm": 0.1137164756655693, "learning_rate": 4.0824893535465434e-05, "loss": 0.0213, "step": 92930 }, { "epoch": 1.8572399184684865, "grad_norm": 0.13514281809329987, "learning_rate": 4.081473502673932e-05, "loss": 0.0313, "step": 92940 }, { "epoch": 1.8574397506094882, "grad_norm": 0.1271735280752182, "learning_rate": 4.080457691039086e-05, "loss": 0.0305, "step": 92950 }, { "epoch": 1.8576395827504895, "grad_norm": 0.1343878209590912, "learning_rate": 4.079441918685399e-05, "loss": 0.0356, "step": 92960 }, { "epoch": 1.8578394148914912, "grad_norm": 0.12499498575925827, "learning_rate": 4.0784261856562625e-05, "loss": 0.0294, "step": 92970 }, { "epoch": 1.8580392470324927, "grad_norm": 0.10509456694126129, "learning_rate": 4.0774104919950636e-05, "loss": 0.0248, "step": 92980 }, { "epoch": 1.8582390791734942, "grad_norm": 0.11670272052288055, "learning_rate": 4.076394837745196e-05, "loss": 0.0275, "step": 92990 }, { "epoch": 1.858438911314496, "grad_norm": 0.12130330502986908, "learning_rate": 4.075379222950044e-05, "loss": 0.0268, "step": 93000 }, { "epoch": 1.8586387434554974, "grad_norm": 0.1038588359951973, "learning_rate": 4.074363647652993e-05, "loss": 0.0321, "step": 93010 }, { "epoch": 1.858838575596499, "grad_norm": 0.1326013207435608, "learning_rate": 4.073348111897427e-05, "loss": 0.0262, "step": 93020 }, { "epoch": 1.8590384077375006, "grad_norm": 0.11698925495147705, "learning_rate": 4.0723326157267253e-05, "loss": 0.0318, "step": 93030 }, { "epoch": 1.859238239878502, "grad_norm": 0.11942120641469955, "learning_rate": 4.071317159184271e-05, "loss": 0.0318, "step": 93040 }, { "epoch": 1.8594380720195036, "grad_norm": 0.14191675186157227, "learning_rate": 4.0703017423134434e-05, "loss": 0.0297, "step": 93050 }, { "epoch": 1.8596379041605053, "grad_norm": 0.1301567107439041, "learning_rate": 4.069286365157616e-05, "loss": 0.0328, "step": 93060 }, { "epoch": 1.8598377363015066, "grad_norm": 0.10495028644800186, "learning_rate": 4.068271027760164e-05, "loss": 0.0253, "step": 93070 }, { "epoch": 1.8600375684425083, "grad_norm": 0.11609438806772232, "learning_rate": 4.0672557301644645e-05, "loss": 0.024, "step": 93080 }, { "epoch": 1.8602374005835098, "grad_norm": 0.11361358314752579, "learning_rate": 4.066240472413886e-05, "loss": 0.0222, "step": 93090 }, { "epoch": 1.8604372327245113, "grad_norm": 0.12047281861305237, "learning_rate": 4.0652252545517996e-05, "loss": 0.0309, "step": 93100 }, { "epoch": 1.860637064865513, "grad_norm": 0.11417003720998764, "learning_rate": 4.064210076621572e-05, "loss": 0.0278, "step": 93110 }, { "epoch": 1.8608368970065146, "grad_norm": 0.13918310403823853, "learning_rate": 4.0631949386665696e-05, "loss": 0.0317, "step": 93120 }, { "epoch": 1.861036729147516, "grad_norm": 0.11594021320343018, "learning_rate": 4.0621798407301606e-05, "loss": 0.0274, "step": 93130 }, { "epoch": 1.8612365612885178, "grad_norm": 0.10242962092161179, "learning_rate": 4.0611647828557056e-05, "loss": 0.0292, "step": 93140 }, { "epoch": 1.861436393429519, "grad_norm": 0.11636893451213837, "learning_rate": 4.060149765086567e-05, "loss": 0.0296, "step": 93150 }, { "epoch": 1.8616362255705208, "grad_norm": 0.11030803620815277, "learning_rate": 4.0591347874661015e-05, "loss": 0.0223, "step": 93160 }, { "epoch": 1.8618360577115223, "grad_norm": 0.12816376984119415, "learning_rate": 4.058119850037671e-05, "loss": 0.0342, "step": 93170 }, { "epoch": 1.8620358898525238, "grad_norm": 0.10610780864953995, "learning_rate": 4.057104952844631e-05, "loss": 0.02, "step": 93180 }, { "epoch": 1.8622357219935255, "grad_norm": 0.11559233069419861, "learning_rate": 4.056090095930334e-05, "loss": 0.0276, "step": 93190 }, { "epoch": 1.862435554134527, "grad_norm": 0.14274659752845764, "learning_rate": 4.055075279338136e-05, "loss": 0.026, "step": 93200 }, { "epoch": 1.8626353862755285, "grad_norm": 0.13188940286636353, "learning_rate": 4.0540605031113834e-05, "loss": 0.0248, "step": 93210 }, { "epoch": 1.8628352184165302, "grad_norm": 0.12269217520952225, "learning_rate": 4.053045767293429e-05, "loss": 0.0343, "step": 93220 }, { "epoch": 1.8630350505575317, "grad_norm": 0.12558241188526154, "learning_rate": 4.0520310719276214e-05, "loss": 0.0322, "step": 93230 }, { "epoch": 1.8632348826985332, "grad_norm": 0.162201926112175, "learning_rate": 4.051016417057304e-05, "loss": 0.0244, "step": 93240 }, { "epoch": 1.863434714839535, "grad_norm": 0.1187170222401619, "learning_rate": 4.050001802725821e-05, "loss": 0.0252, "step": 93250 }, { "epoch": 1.8636345469805362, "grad_norm": 0.10425388813018799, "learning_rate": 4.048987228976518e-05, "loss": 0.021, "step": 93260 }, { "epoch": 1.863834379121538, "grad_norm": 0.1017071008682251, "learning_rate": 4.0479726958527314e-05, "loss": 0.0294, "step": 93270 }, { "epoch": 1.8640342112625394, "grad_norm": 0.13124485313892365, "learning_rate": 4.0469582033978036e-05, "loss": 0.0298, "step": 93280 }, { "epoch": 1.864234043403541, "grad_norm": 0.15408460795879364, "learning_rate": 4.0459437516550705e-05, "loss": 0.03, "step": 93290 }, { "epoch": 1.8644338755445427, "grad_norm": 0.13111494481563568, "learning_rate": 4.044929340667866e-05, "loss": 0.0263, "step": 93300 }, { "epoch": 1.8646337076855442, "grad_norm": 0.11999927461147308, "learning_rate": 4.0439149704795276e-05, "loss": 0.0247, "step": 93310 }, { "epoch": 1.8648335398265456, "grad_norm": 0.15267314016819, "learning_rate": 4.042900641133384e-05, "loss": 0.0234, "step": 93320 }, { "epoch": 1.8650333719675474, "grad_norm": 0.11015865951776505, "learning_rate": 4.041886352672769e-05, "loss": 0.0348, "step": 93330 }, { "epoch": 1.8652332041085489, "grad_norm": 0.11235212534666061, "learning_rate": 4.040872105141005e-05, "loss": 0.033, "step": 93340 }, { "epoch": 1.8654330362495504, "grad_norm": 0.11226674914360046, "learning_rate": 4.039857898581425e-05, "loss": 0.0254, "step": 93350 }, { "epoch": 1.865632868390552, "grad_norm": 0.1473890244960785, "learning_rate": 4.038843733037352e-05, "loss": 0.0322, "step": 93360 }, { "epoch": 1.8658327005315534, "grad_norm": 0.11341797560453415, "learning_rate": 4.037829608552109e-05, "loss": 0.0268, "step": 93370 }, { "epoch": 1.866032532672555, "grad_norm": 0.11977341771125793, "learning_rate": 4.036815525169018e-05, "loss": 0.0287, "step": 93380 }, { "epoch": 1.8662323648135566, "grad_norm": 0.1445358246564865, "learning_rate": 4.0358014829313954e-05, "loss": 0.0243, "step": 93390 }, { "epoch": 1.866432196954558, "grad_norm": 0.09984851628541946, "learning_rate": 4.034787481882565e-05, "loss": 0.0261, "step": 93400 }, { "epoch": 1.8666320290955598, "grad_norm": 0.21082304418087006, "learning_rate": 4.033773522065839e-05, "loss": 0.0345, "step": 93410 }, { "epoch": 1.8668318612365613, "grad_norm": 0.10882224142551422, "learning_rate": 4.0327596035245345e-05, "loss": 0.0258, "step": 93420 }, { "epoch": 1.8670316933775628, "grad_norm": 0.10714547336101532, "learning_rate": 4.031745726301961e-05, "loss": 0.03, "step": 93430 }, { "epoch": 1.8672315255185645, "grad_norm": 0.13867439329624176, "learning_rate": 4.030731890441432e-05, "loss": 0.0225, "step": 93440 }, { "epoch": 1.8674313576595658, "grad_norm": 0.14404477179050446, "learning_rate": 4.029718095986258e-05, "loss": 0.0313, "step": 93450 }, { "epoch": 1.8676311898005675, "grad_norm": 0.11443416774272919, "learning_rate": 4.028704342979742e-05, "loss": 0.0259, "step": 93460 }, { "epoch": 1.8678310219415692, "grad_norm": 0.11572784185409546, "learning_rate": 4.027690631465192e-05, "loss": 0.0364, "step": 93470 }, { "epoch": 1.8680308540825705, "grad_norm": 0.12768293917179108, "learning_rate": 4.026676961485914e-05, "loss": 0.0232, "step": 93480 }, { "epoch": 1.8682306862235722, "grad_norm": 0.15576477348804474, "learning_rate": 4.0256633330852066e-05, "loss": 0.0281, "step": 93490 }, { "epoch": 1.8684305183645737, "grad_norm": 0.19429637491703033, "learning_rate": 4.024649746306373e-05, "loss": 0.0344, "step": 93500 }, { "epoch": 1.8686303505055752, "grad_norm": 0.14976361393928528, "learning_rate": 4.02363620119271e-05, "loss": 0.0238, "step": 93510 }, { "epoch": 1.868830182646577, "grad_norm": 0.092367023229599, "learning_rate": 4.0226226977875125e-05, "loss": 0.0275, "step": 93520 }, { "epoch": 1.8690300147875785, "grad_norm": 0.10314077883958817, "learning_rate": 4.0216092361340814e-05, "loss": 0.0313, "step": 93530 }, { "epoch": 1.86922984692858, "grad_norm": 0.09251224994659424, "learning_rate": 4.020595816275705e-05, "loss": 0.0268, "step": 93540 }, { "epoch": 1.8694296790695817, "grad_norm": 0.10436616092920303, "learning_rate": 4.0195824382556764e-05, "loss": 0.0278, "step": 93550 }, { "epoch": 1.869629511210583, "grad_norm": 0.14200495183467865, "learning_rate": 4.0185691021172835e-05, "loss": 0.0281, "step": 93560 }, { "epoch": 1.8698293433515847, "grad_norm": 0.1321970373392105, "learning_rate": 4.017555807903816e-05, "loss": 0.0265, "step": 93570 }, { "epoch": 1.8700291754925862, "grad_norm": 0.12454046308994293, "learning_rate": 4.016542555658561e-05, "loss": 0.0235, "step": 93580 }, { "epoch": 1.8702290076335877, "grad_norm": 0.10836930572986603, "learning_rate": 4.0155293454248014e-05, "loss": 0.025, "step": 93590 }, { "epoch": 1.8704288397745894, "grad_norm": 0.1248743087053299, "learning_rate": 4.014516177245819e-05, "loss": 0.0317, "step": 93600 }, { "epoch": 1.870628671915591, "grad_norm": 0.13419362902641296, "learning_rate": 4.0135030511648934e-05, "loss": 0.0304, "step": 93610 }, { "epoch": 1.8708285040565924, "grad_norm": 0.09528923779726028, "learning_rate": 4.012489967225307e-05, "loss": 0.0327, "step": 93620 }, { "epoch": 1.8710283361975941, "grad_norm": 0.09951181709766388, "learning_rate": 4.0114769254703356e-05, "loss": 0.027, "step": 93630 }, { "epoch": 1.8712281683385956, "grad_norm": 0.09261748194694519, "learning_rate": 4.010463925943253e-05, "loss": 0.0227, "step": 93640 }, { "epoch": 1.8714280004795971, "grad_norm": 0.12405430525541306, "learning_rate": 4.009450968687333e-05, "loss": 0.0285, "step": 93650 }, { "epoch": 1.8716278326205988, "grad_norm": 0.11959081888198853, "learning_rate": 4.0084380537458496e-05, "loss": 0.0357, "step": 93660 }, { "epoch": 1.8718276647616001, "grad_norm": 0.11870954930782318, "learning_rate": 4.007425181162069e-05, "loss": 0.0353, "step": 93670 }, { "epoch": 1.8720274969026018, "grad_norm": 0.11963728815317154, "learning_rate": 4.006412350979264e-05, "loss": 0.0233, "step": 93680 }, { "epoch": 1.8722273290436033, "grad_norm": 0.11261474341154099, "learning_rate": 4.005399563240696e-05, "loss": 0.0256, "step": 93690 }, { "epoch": 1.8724271611846048, "grad_norm": 0.1370212435722351, "learning_rate": 4.004386817989631e-05, "loss": 0.032, "step": 93700 }, { "epoch": 1.8726269933256066, "grad_norm": 0.15800319612026215, "learning_rate": 4.0033741152693337e-05, "loss": 0.0343, "step": 93710 }, { "epoch": 1.872826825466608, "grad_norm": 0.10430081188678741, "learning_rate": 4.002361455123062e-05, "loss": 0.024, "step": 93720 }, { "epoch": 1.8730266576076096, "grad_norm": 0.16769655048847198, "learning_rate": 4.001348837594079e-05, "loss": 0.0288, "step": 93730 }, { "epoch": 1.8732264897486113, "grad_norm": 0.1076204776763916, "learning_rate": 4.000336262725635e-05, "loss": 0.0227, "step": 93740 }, { "epoch": 1.8734263218896128, "grad_norm": 0.1285482496023178, "learning_rate": 3.999323730560992e-05, "loss": 0.0298, "step": 93750 }, { "epoch": 1.8736261540306143, "grad_norm": 0.12461210787296295, "learning_rate": 3.998311241143401e-05, "loss": 0.0317, "step": 93760 }, { "epoch": 1.873825986171616, "grad_norm": 0.09595663845539093, "learning_rate": 3.997298794516113e-05, "loss": 0.0239, "step": 93770 }, { "epoch": 1.8740258183126173, "grad_norm": 0.13832026720046997, "learning_rate": 3.996286390722379e-05, "loss": 0.0207, "step": 93780 }, { "epoch": 1.874225650453619, "grad_norm": 0.1406945139169693, "learning_rate": 3.995274029805444e-05, "loss": 0.0246, "step": 93790 }, { "epoch": 1.8744254825946205, "grad_norm": 0.09916974604129791, "learning_rate": 3.9942617118085594e-05, "loss": 0.0285, "step": 93800 }, { "epoch": 1.874625314735622, "grad_norm": 0.16157637536525726, "learning_rate": 3.993249436774966e-05, "loss": 0.0233, "step": 93810 }, { "epoch": 1.8748251468766237, "grad_norm": 0.1123991310596466, "learning_rate": 3.992237204747908e-05, "loss": 0.0302, "step": 93820 }, { "epoch": 1.8750249790176252, "grad_norm": 0.1124197393655777, "learning_rate": 3.9912250157706236e-05, "loss": 0.0309, "step": 93830 }, { "epoch": 1.8752248111586267, "grad_norm": 0.13961005210876465, "learning_rate": 3.990212869886354e-05, "loss": 0.0295, "step": 93840 }, { "epoch": 1.8754246432996284, "grad_norm": 0.186768040060997, "learning_rate": 3.9892007671383375e-05, "loss": 0.0314, "step": 93850 }, { "epoch": 1.8756244754406297, "grad_norm": 0.13795709609985352, "learning_rate": 3.988188707569805e-05, "loss": 0.0326, "step": 93860 }, { "epoch": 1.8758243075816314, "grad_norm": 0.11780072003602982, "learning_rate": 3.987176691223994e-05, "loss": 0.0312, "step": 93870 }, { "epoch": 1.876024139722633, "grad_norm": 0.09136256575584412, "learning_rate": 3.986164718144132e-05, "loss": 0.0349, "step": 93880 }, { "epoch": 1.8762239718636344, "grad_norm": 0.20207689702510834, "learning_rate": 3.9851527883734516e-05, "loss": 0.0363, "step": 93890 }, { "epoch": 1.8764238040046362, "grad_norm": 0.1381765902042389, "learning_rate": 3.984140901955181e-05, "loss": 0.0307, "step": 93900 }, { "epoch": 1.8766236361456377, "grad_norm": 0.10058339685201645, "learning_rate": 3.9831290589325435e-05, "loss": 0.03, "step": 93910 }, { "epoch": 1.8768234682866392, "grad_norm": 0.10123614221811295, "learning_rate": 3.982117259348764e-05, "loss": 0.0263, "step": 93920 }, { "epoch": 1.8770233004276409, "grad_norm": 0.09492282569408417, "learning_rate": 3.981105503247067e-05, "loss": 0.0185, "step": 93930 }, { "epoch": 1.8772231325686424, "grad_norm": 0.11875753849744797, "learning_rate": 3.9800937906706705e-05, "loss": 0.0273, "step": 93940 }, { "epoch": 1.8774229647096439, "grad_norm": 0.1497311145067215, "learning_rate": 3.979082121662794e-05, "loss": 0.0243, "step": 93950 }, { "epoch": 1.8776227968506456, "grad_norm": 0.11689439415931702, "learning_rate": 3.978070496266653e-05, "loss": 0.0258, "step": 93960 }, { "epoch": 1.8778226289916469, "grad_norm": 0.1181645616889, "learning_rate": 3.977058914525462e-05, "loss": 0.0294, "step": 93970 }, { "epoch": 1.8780224611326486, "grad_norm": 0.12074673175811768, "learning_rate": 3.9760473764824374e-05, "loss": 0.0285, "step": 93980 }, { "epoch": 1.87822229327365, "grad_norm": 0.12873157858848572, "learning_rate": 3.975035882180786e-05, "loss": 0.0311, "step": 93990 }, { "epoch": 1.8784221254146516, "grad_norm": 0.12772347033023834, "learning_rate": 3.974024431663721e-05, "loss": 0.0302, "step": 94000 }, { "epoch": 1.8786219575556533, "grad_norm": 0.17601267993450165, "learning_rate": 3.973013024974445e-05, "loss": 0.0242, "step": 94010 }, { "epoch": 1.8788217896966548, "grad_norm": 0.1753072887659073, "learning_rate": 3.972001662156166e-05, "loss": 0.0277, "step": 94020 }, { "epoch": 1.8790216218376563, "grad_norm": 0.17609187960624695, "learning_rate": 3.97099034325209e-05, "loss": 0.0411, "step": 94030 }, { "epoch": 1.879221453978658, "grad_norm": 0.10087359696626663, "learning_rate": 3.969979068305414e-05, "loss": 0.0185, "step": 94040 }, { "epoch": 1.8794212861196595, "grad_norm": 0.1308603286743164, "learning_rate": 3.968967837359342e-05, "loss": 0.0303, "step": 94050 }, { "epoch": 1.879621118260661, "grad_norm": 0.12916257977485657, "learning_rate": 3.967956650457067e-05, "loss": 0.0324, "step": 94060 }, { "epoch": 1.8798209504016627, "grad_norm": 0.12876597046852112, "learning_rate": 3.96694550764179e-05, "loss": 0.0332, "step": 94070 }, { "epoch": 1.880020782542664, "grad_norm": 0.1259436160326004, "learning_rate": 3.965934408956703e-05, "loss": 0.0285, "step": 94080 }, { "epoch": 1.8802206146836657, "grad_norm": 0.14857029914855957, "learning_rate": 3.964923354444997e-05, "loss": 0.0239, "step": 94090 }, { "epoch": 1.8804204468246672, "grad_norm": 0.1265912503004074, "learning_rate": 3.9639123441498645e-05, "loss": 0.0389, "step": 94100 }, { "epoch": 1.8806202789656687, "grad_norm": 0.15286129713058472, "learning_rate": 3.962901378114494e-05, "loss": 0.0275, "step": 94110 }, { "epoch": 1.8808201111066705, "grad_norm": 0.13051584362983704, "learning_rate": 3.961890456382071e-05, "loss": 0.0248, "step": 94120 }, { "epoch": 1.881019943247672, "grad_norm": 0.12218929082155228, "learning_rate": 3.960879578995782e-05, "loss": 0.034, "step": 94130 }, { "epoch": 1.8812197753886735, "grad_norm": 0.12875337898731232, "learning_rate": 3.959868745998807e-05, "loss": 0.0296, "step": 94140 }, { "epoch": 1.8814196075296752, "grad_norm": 0.17667128145694733, "learning_rate": 3.958857957434327e-05, "loss": 0.0242, "step": 94150 }, { "epoch": 1.8816194396706767, "grad_norm": 0.14921097457408905, "learning_rate": 3.957847213345525e-05, "loss": 0.0415, "step": 94160 }, { "epoch": 1.8818192718116782, "grad_norm": 0.13479582965373993, "learning_rate": 3.956836513775574e-05, "loss": 0.0252, "step": 94170 }, { "epoch": 1.88201910395268, "grad_norm": 0.16988453269004822, "learning_rate": 3.955825858767652e-05, "loss": 0.0398, "step": 94180 }, { "epoch": 1.8822189360936812, "grad_norm": 0.12577320635318756, "learning_rate": 3.954815248364928e-05, "loss": 0.0309, "step": 94190 }, { "epoch": 1.882418768234683, "grad_norm": 0.1273166984319687, "learning_rate": 3.95380468261058e-05, "loss": 0.0214, "step": 94200 }, { "epoch": 1.8826186003756844, "grad_norm": 0.1487167924642563, "learning_rate": 3.952794161547773e-05, "loss": 0.0255, "step": 94210 }, { "epoch": 1.882818432516686, "grad_norm": 0.11921212822198868, "learning_rate": 3.9517836852196767e-05, "loss": 0.0393, "step": 94220 }, { "epoch": 1.8830182646576876, "grad_norm": 0.11960260570049286, "learning_rate": 3.9507732536694545e-05, "loss": 0.028, "step": 94230 }, { "epoch": 1.8832180967986891, "grad_norm": 0.18538908660411835, "learning_rate": 3.94976286694027e-05, "loss": 0.0292, "step": 94240 }, { "epoch": 1.8834179289396906, "grad_norm": 0.17186181247234344, "learning_rate": 3.948752525075289e-05, "loss": 0.0356, "step": 94250 }, { "epoch": 1.8836177610806923, "grad_norm": 0.13474112749099731, "learning_rate": 3.947742228117668e-05, "loss": 0.0272, "step": 94260 }, { "epoch": 1.8838175932216936, "grad_norm": 0.11978907138109207, "learning_rate": 3.946731976110566e-05, "loss": 0.0255, "step": 94270 }, { "epoch": 1.8840174253626953, "grad_norm": 0.12342256307601929, "learning_rate": 3.945721769097138e-05, "loss": 0.0278, "step": 94280 }, { "epoch": 1.8842172575036968, "grad_norm": 0.14858126640319824, "learning_rate": 3.9447116071205406e-05, "loss": 0.0248, "step": 94290 }, { "epoch": 1.8844170896446983, "grad_norm": 0.08177207410335541, "learning_rate": 3.943701490223925e-05, "loss": 0.0229, "step": 94300 }, { "epoch": 1.8846169217857, "grad_norm": 0.11342236399650574, "learning_rate": 3.94269141845044e-05, "loss": 0.0302, "step": 94310 }, { "epoch": 1.8848167539267016, "grad_norm": 0.12148884683847427, "learning_rate": 3.941681391843237e-05, "loss": 0.0356, "step": 94320 }, { "epoch": 1.885016586067703, "grad_norm": 0.12552668154239655, "learning_rate": 3.940671410445459e-05, "loss": 0.0284, "step": 94330 }, { "epoch": 1.8852164182087048, "grad_norm": 0.10043150186538696, "learning_rate": 3.939661474300252e-05, "loss": 0.0251, "step": 94340 }, { "epoch": 1.8854162503497063, "grad_norm": 0.12104425579309464, "learning_rate": 3.938651583450761e-05, "loss": 0.0305, "step": 94350 }, { "epoch": 1.8856160824907078, "grad_norm": 0.09880097955465317, "learning_rate": 3.9376417379401234e-05, "loss": 0.0254, "step": 94360 }, { "epoch": 1.8858159146317095, "grad_norm": 0.10614216327667236, "learning_rate": 3.936631937811478e-05, "loss": 0.032, "step": 94370 }, { "epoch": 1.8860157467727108, "grad_norm": 0.10995586961507797, "learning_rate": 3.935622183107964e-05, "loss": 0.0294, "step": 94380 }, { "epoch": 1.8862155789137125, "grad_norm": 0.11158201098442078, "learning_rate": 3.934612473872715e-05, "loss": 0.0236, "step": 94390 }, { "epoch": 1.886415411054714, "grad_norm": 0.11453491449356079, "learning_rate": 3.933602810148865e-05, "loss": 0.0265, "step": 94400 }, { "epoch": 1.8866152431957155, "grad_norm": 0.08874082565307617, "learning_rate": 3.932593191979542e-05, "loss": 0.026, "step": 94410 }, { "epoch": 1.8868150753367172, "grad_norm": 0.10449646413326263, "learning_rate": 3.931583619407877e-05, "loss": 0.0286, "step": 94420 }, { "epoch": 1.8870149074777187, "grad_norm": 0.1411903351545334, "learning_rate": 3.930574092476999e-05, "loss": 0.023, "step": 94430 }, { "epoch": 1.8872147396187202, "grad_norm": 0.13127417862415314, "learning_rate": 3.929564611230029e-05, "loss": 0.0275, "step": 94440 }, { "epoch": 1.887414571759722, "grad_norm": 0.10776550322771072, "learning_rate": 3.928555175710095e-05, "loss": 0.0346, "step": 94450 }, { "epoch": 1.8876144039007234, "grad_norm": 0.11785823851823807, "learning_rate": 3.927545785960312e-05, "loss": 0.0335, "step": 94460 }, { "epoch": 1.887814236041725, "grad_norm": 0.12347155809402466, "learning_rate": 3.926536442023804e-05, "loss": 0.0267, "step": 94470 }, { "epoch": 1.8880140681827267, "grad_norm": 0.12958841025829315, "learning_rate": 3.925527143943689e-05, "loss": 0.0267, "step": 94480 }, { "epoch": 1.888213900323728, "grad_norm": 0.10578135401010513, "learning_rate": 3.92451789176308e-05, "loss": 0.0217, "step": 94490 }, { "epoch": 1.8884137324647297, "grad_norm": 0.111679308116436, "learning_rate": 3.923508685525091e-05, "loss": 0.0364, "step": 94500 }, { "epoch": 1.8886135646057312, "grad_norm": 0.14476943016052246, "learning_rate": 3.922499525272832e-05, "loss": 0.0252, "step": 94510 }, { "epoch": 1.8888133967467327, "grad_norm": 0.16578824818134308, "learning_rate": 3.921490411049414e-05, "loss": 0.0272, "step": 94520 }, { "epoch": 1.8890132288877344, "grad_norm": 0.08685926347970963, "learning_rate": 3.920481342897946e-05, "loss": 0.0251, "step": 94530 }, { "epoch": 1.8892130610287359, "grad_norm": 0.08603590726852417, "learning_rate": 3.91947232086153e-05, "loss": 0.0228, "step": 94540 }, { "epoch": 1.8894128931697374, "grad_norm": 0.10261348634958267, "learning_rate": 3.918463344983271e-05, "loss": 0.0336, "step": 94550 }, { "epoch": 1.889612725310739, "grad_norm": 0.16840332746505737, "learning_rate": 3.917454415306274e-05, "loss": 0.0233, "step": 94560 }, { "epoch": 1.8898125574517406, "grad_norm": 0.13255572319030762, "learning_rate": 3.9164455318736334e-05, "loss": 0.0277, "step": 94570 }, { "epoch": 1.890012389592742, "grad_norm": 0.11217144131660461, "learning_rate": 3.915436694728451e-05, "loss": 0.0274, "step": 94580 }, { "epoch": 1.8902122217337438, "grad_norm": 0.15183207392692566, "learning_rate": 3.914427903913817e-05, "loss": 0.0284, "step": 94590 }, { "epoch": 1.890412053874745, "grad_norm": 0.10299552977085114, "learning_rate": 3.913419159472832e-05, "loss": 0.0343, "step": 94600 }, { "epoch": 1.8906118860157468, "grad_norm": 0.12106884270906448, "learning_rate": 3.912410461448583e-05, "loss": 0.0316, "step": 94610 }, { "epoch": 1.8908117181567483, "grad_norm": 0.14973679184913635, "learning_rate": 3.9114018098841624e-05, "loss": 0.032, "step": 94620 }, { "epoch": 1.8910115502977498, "grad_norm": 0.14708212018013, "learning_rate": 3.910393204822656e-05, "loss": 0.0334, "step": 94630 }, { "epoch": 1.8912113824387515, "grad_norm": 0.15034489333629608, "learning_rate": 3.909384646307148e-05, "loss": 0.0369, "step": 94640 }, { "epoch": 1.891411214579753, "grad_norm": 0.1493138074874878, "learning_rate": 3.908376134380728e-05, "loss": 0.0351, "step": 94650 }, { "epoch": 1.8916110467207545, "grad_norm": 0.12690357863903046, "learning_rate": 3.907367669086472e-05, "loss": 0.0244, "step": 94660 }, { "epoch": 1.8918108788617563, "grad_norm": 0.14768998324871063, "learning_rate": 3.906359250467464e-05, "loss": 0.0279, "step": 94670 }, { "epoch": 1.8920107110027575, "grad_norm": 0.13116228580474854, "learning_rate": 3.905350878566777e-05, "loss": 0.0288, "step": 94680 }, { "epoch": 1.8922105431437592, "grad_norm": 0.12239717692136765, "learning_rate": 3.90434255342749e-05, "loss": 0.0216, "step": 94690 }, { "epoch": 1.8924103752847607, "grad_norm": 0.16124069690704346, "learning_rate": 3.903334275092678e-05, "loss": 0.0215, "step": 94700 }, { "epoch": 1.8926102074257622, "grad_norm": 0.11173523217439651, "learning_rate": 3.9023260436054106e-05, "loss": 0.0267, "step": 94710 }, { "epoch": 1.892810039566764, "grad_norm": 0.12692119181156158, "learning_rate": 3.9013178590087584e-05, "loss": 0.0263, "step": 94720 }, { "epoch": 1.8930098717077655, "grad_norm": 0.16459062695503235, "learning_rate": 3.900309721345787e-05, "loss": 0.0282, "step": 94730 }, { "epoch": 1.893209703848767, "grad_norm": 0.13429316878318787, "learning_rate": 3.899301630659565e-05, "loss": 0.0216, "step": 94740 }, { "epoch": 1.8934095359897687, "grad_norm": 0.15595999360084534, "learning_rate": 3.8982935869931566e-05, "loss": 0.0295, "step": 94750 }, { "epoch": 1.8936093681307702, "grad_norm": 0.13676349818706512, "learning_rate": 3.897285590389622e-05, "loss": 0.0286, "step": 94760 }, { "epoch": 1.8938092002717717, "grad_norm": 0.1257997453212738, "learning_rate": 3.896277640892019e-05, "loss": 0.0248, "step": 94770 }, { "epoch": 1.8940090324127734, "grad_norm": 0.21633169054985046, "learning_rate": 3.89526973854341e-05, "loss": 0.0263, "step": 94780 }, { "epoch": 1.8942088645537747, "grad_norm": 0.09569089859724045, "learning_rate": 3.8942618833868474e-05, "loss": 0.0293, "step": 94790 }, { "epoch": 1.8944086966947764, "grad_norm": 0.11581265181303024, "learning_rate": 3.893254075465387e-05, "loss": 0.024, "step": 94800 }, { "epoch": 1.894608528835778, "grad_norm": 0.13290974497795105, "learning_rate": 3.8922463148220775e-05, "loss": 0.0279, "step": 94810 }, { "epoch": 1.8948083609767794, "grad_norm": 0.11100441962480545, "learning_rate": 3.8912386014999705e-05, "loss": 0.0226, "step": 94820 }, { "epoch": 1.8950081931177811, "grad_norm": 0.12879642844200134, "learning_rate": 3.890230935542115e-05, "loss": 0.0314, "step": 94830 }, { "epoch": 1.8952080252587826, "grad_norm": 0.08761081844568253, "learning_rate": 3.889223316991555e-05, "loss": 0.0332, "step": 94840 }, { "epoch": 1.8954078573997841, "grad_norm": 0.10836672782897949, "learning_rate": 3.8882157458913344e-05, "loss": 0.0322, "step": 94850 }, { "epoch": 1.8956076895407858, "grad_norm": 0.14664234220981598, "learning_rate": 3.887208222284493e-05, "loss": 0.0332, "step": 94860 }, { "epoch": 1.8958075216817873, "grad_norm": 0.09768076241016388, "learning_rate": 3.886200746214072e-05, "loss": 0.0225, "step": 94870 }, { "epoch": 1.8960073538227888, "grad_norm": 0.1248324066400528, "learning_rate": 3.885193317723112e-05, "loss": 0.0211, "step": 94880 }, { "epoch": 1.8962071859637906, "grad_norm": 0.13612227141857147, "learning_rate": 3.8841859368546434e-05, "loss": 0.0261, "step": 94890 }, { "epoch": 1.8964070181047918, "grad_norm": 0.09106134623289108, "learning_rate": 3.883178603651703e-05, "loss": 0.025, "step": 94900 }, { "epoch": 1.8966068502457936, "grad_norm": 0.1424614042043686, "learning_rate": 3.882171318157319e-05, "loss": 0.0315, "step": 94910 }, { "epoch": 1.896806682386795, "grad_norm": 0.14475037157535553, "learning_rate": 3.881164080414524e-05, "loss": 0.0293, "step": 94920 }, { "epoch": 1.8970065145277966, "grad_norm": 0.12254932522773743, "learning_rate": 3.8801568904663447e-05, "loss": 0.032, "step": 94930 }, { "epoch": 1.8972063466687983, "grad_norm": 0.15391646325588226, "learning_rate": 3.8791497483558046e-05, "loss": 0.0264, "step": 94940 }, { "epoch": 1.8974061788097998, "grad_norm": 0.10528619587421417, "learning_rate": 3.8781426541259283e-05, "loss": 0.0281, "step": 94950 }, { "epoch": 1.8976060109508013, "grad_norm": 0.09888801723718643, "learning_rate": 3.877135607819736e-05, "loss": 0.0255, "step": 94960 }, { "epoch": 1.897805843091803, "grad_norm": 0.12149731069803238, "learning_rate": 3.876128609480251e-05, "loss": 0.0355, "step": 94970 }, { "epoch": 1.8980056752328043, "grad_norm": 0.12188570946455002, "learning_rate": 3.8751216591504845e-05, "loss": 0.0289, "step": 94980 }, { "epoch": 1.898205507373806, "grad_norm": 0.10722916573286057, "learning_rate": 3.8741147568734556e-05, "loss": 0.0231, "step": 94990 }, { "epoch": 1.8984053395148077, "grad_norm": 0.1130756139755249, "learning_rate": 3.8731079026921745e-05, "loss": 0.0323, "step": 95000 }, { "epoch": 1.898605171655809, "grad_norm": 0.14060528576374054, "learning_rate": 3.872101096649654e-05, "loss": 0.0305, "step": 95010 }, { "epoch": 1.8988050037968107, "grad_norm": 0.14315663278102875, "learning_rate": 3.871094338788904e-05, "loss": 0.0219, "step": 95020 }, { "epoch": 1.8990048359378122, "grad_norm": 0.11180508881807327, "learning_rate": 3.8700876291529294e-05, "loss": 0.0308, "step": 95030 }, { "epoch": 1.8992046680788137, "grad_norm": 0.1304183304309845, "learning_rate": 3.869080967784734e-05, "loss": 0.0308, "step": 95040 }, { "epoch": 1.8994045002198154, "grad_norm": 0.12430873513221741, "learning_rate": 3.868074354727324e-05, "loss": 0.0206, "step": 95050 }, { "epoch": 1.899604332360817, "grad_norm": 0.1250084787607193, "learning_rate": 3.8670677900236977e-05, "loss": 0.0236, "step": 95060 }, { "epoch": 1.8998041645018184, "grad_norm": 0.09441389888525009, "learning_rate": 3.8660612737168546e-05, "loss": 0.0288, "step": 95070 }, { "epoch": 1.9000039966428202, "grad_norm": 0.15845859050750732, "learning_rate": 3.86505480584979e-05, "loss": 0.0321, "step": 95080 }, { "epoch": 1.9002038287838214, "grad_norm": 0.1400781273841858, "learning_rate": 3.8640483864654984e-05, "loss": 0.0312, "step": 95090 }, { "epoch": 1.9004036609248232, "grad_norm": 0.13335764408111572, "learning_rate": 3.863042015606975e-05, "loss": 0.0321, "step": 95100 }, { "epoch": 1.9006034930658247, "grad_norm": 0.1494237184524536, "learning_rate": 3.862035693317206e-05, "loss": 0.0238, "step": 95110 }, { "epoch": 1.9008033252068262, "grad_norm": 0.14849025011062622, "learning_rate": 3.8610294196391835e-05, "loss": 0.0281, "step": 95120 }, { "epoch": 1.9010031573478279, "grad_norm": 0.1397620439529419, "learning_rate": 3.860023194615889e-05, "loss": 0.03, "step": 95130 }, { "epoch": 1.9012029894888294, "grad_norm": 0.09454022347927094, "learning_rate": 3.859017018290312e-05, "loss": 0.0312, "step": 95140 }, { "epoch": 1.9014028216298309, "grad_norm": 0.1512393057346344, "learning_rate": 3.858010890705432e-05, "loss": 0.0256, "step": 95150 }, { "epoch": 1.9016026537708326, "grad_norm": 0.1244281679391861, "learning_rate": 3.857004811904227e-05, "loss": 0.0238, "step": 95160 }, { "epoch": 1.901802485911834, "grad_norm": 0.09834518283605576, "learning_rate": 3.8559987819296794e-05, "loss": 0.0204, "step": 95170 }, { "epoch": 1.9020023180528356, "grad_norm": 0.19037388265132904, "learning_rate": 3.854992800824759e-05, "loss": 0.031, "step": 95180 }, { "epoch": 1.9022021501938373, "grad_norm": 0.13509075343608856, "learning_rate": 3.8539868686324443e-05, "loss": 0.0334, "step": 95190 }, { "epoch": 1.9024019823348386, "grad_norm": 0.11170786619186401, "learning_rate": 3.852980985395706e-05, "loss": 0.0307, "step": 95200 }, { "epoch": 1.9026018144758403, "grad_norm": 0.12686018645763397, "learning_rate": 3.851975151157512e-05, "loss": 0.0336, "step": 95210 }, { "epoch": 1.9028016466168418, "grad_norm": 0.17461851239204407, "learning_rate": 3.8509693659608296e-05, "loss": 0.0287, "step": 95220 }, { "epoch": 1.9030014787578433, "grad_norm": 0.15653321146965027, "learning_rate": 3.849963629848627e-05, "loss": 0.0322, "step": 95230 }, { "epoch": 1.903201310898845, "grad_norm": 0.11019837111234665, "learning_rate": 3.8489579428638645e-05, "loss": 0.0274, "step": 95240 }, { "epoch": 1.9034011430398465, "grad_norm": 0.12065231055021286, "learning_rate": 3.847952305049505e-05, "loss": 0.0281, "step": 95250 }, { "epoch": 1.903600975180848, "grad_norm": 0.14365942776203156, "learning_rate": 3.846946716448505e-05, "loss": 0.0343, "step": 95260 }, { "epoch": 1.9038008073218498, "grad_norm": 0.10916101932525635, "learning_rate": 3.845941177103823e-05, "loss": 0.0269, "step": 95270 }, { "epoch": 1.9040006394628513, "grad_norm": 0.1316121369600296, "learning_rate": 3.844935687058416e-05, "loss": 0.0284, "step": 95280 }, { "epoch": 1.9042004716038528, "grad_norm": 0.11620084196329117, "learning_rate": 3.843930246355233e-05, "loss": 0.0313, "step": 95290 }, { "epoch": 1.9044003037448545, "grad_norm": 0.12289274483919144, "learning_rate": 3.8429248550372274e-05, "loss": 0.0325, "step": 95300 }, { "epoch": 1.9046001358858557, "grad_norm": 0.1026742160320282, "learning_rate": 3.841919513147344e-05, "loss": 0.0326, "step": 95310 }, { "epoch": 1.9047999680268575, "grad_norm": 0.12354077398777008, "learning_rate": 3.8409142207285326e-05, "loss": 0.0248, "step": 95320 }, { "epoch": 1.904999800167859, "grad_norm": 0.10734020173549652, "learning_rate": 3.839908977823737e-05, "loss": 0.0251, "step": 95330 }, { "epoch": 1.9051996323088605, "grad_norm": 0.0926353707909584, "learning_rate": 3.838903784475899e-05, "loss": 0.0285, "step": 95340 }, { "epoch": 1.9053994644498622, "grad_norm": 0.15396486222743988, "learning_rate": 3.8378986407279585e-05, "loss": 0.0266, "step": 95350 }, { "epoch": 1.9055992965908637, "grad_norm": 0.12715081870555878, "learning_rate": 3.83689354662285e-05, "loss": 0.0291, "step": 95360 }, { "epoch": 1.9057991287318652, "grad_norm": 0.1239141896367073, "learning_rate": 3.835888502203516e-05, "loss": 0.038, "step": 95370 }, { "epoch": 1.905998960872867, "grad_norm": 0.12897029519081116, "learning_rate": 3.834883507512885e-05, "loss": 0.0219, "step": 95380 }, { "epoch": 1.9061987930138682, "grad_norm": 0.12198926508426666, "learning_rate": 3.8338785625938915e-05, "loss": 0.0333, "step": 95390 }, { "epoch": 1.90639862515487, "grad_norm": 0.10270090401172638, "learning_rate": 3.832873667489462e-05, "loss": 0.0319, "step": 95400 }, { "epoch": 1.9065984572958716, "grad_norm": 0.14277151226997375, "learning_rate": 3.831868822242526e-05, "loss": 0.0287, "step": 95410 }, { "epoch": 1.906798289436873, "grad_norm": 0.11110103875398636, "learning_rate": 3.830864026896009e-05, "loss": 0.0323, "step": 95420 }, { "epoch": 1.9069981215778746, "grad_norm": 0.11794336140155792, "learning_rate": 3.8298592814928316e-05, "loss": 0.0303, "step": 95430 }, { "epoch": 1.9071979537188761, "grad_norm": 0.10718692094087601, "learning_rate": 3.828854586075918e-05, "loss": 0.0259, "step": 95440 }, { "epoch": 1.9073977858598776, "grad_norm": 0.1327483057975769, "learning_rate": 3.8278499406881824e-05, "loss": 0.0236, "step": 95450 }, { "epoch": 1.9075976180008793, "grad_norm": 0.1327037662267685, "learning_rate": 3.826845345372545e-05, "loss": 0.0261, "step": 95460 }, { "epoch": 1.9077974501418808, "grad_norm": 0.11508812010288239, "learning_rate": 3.82584080017192e-05, "loss": 0.0306, "step": 95470 }, { "epoch": 1.9079972822828823, "grad_norm": 0.12286963313817978, "learning_rate": 3.824836305129219e-05, "loss": 0.0255, "step": 95480 }, { "epoch": 1.908197114423884, "grad_norm": 0.09193386137485504, "learning_rate": 3.82383186028735e-05, "loss": 0.0223, "step": 95490 }, { "epoch": 1.9083969465648853, "grad_norm": 0.12728005647659302, "learning_rate": 3.822827465689226e-05, "loss": 0.0271, "step": 95500 }, { "epoch": 1.908596778705887, "grad_norm": 0.1144898310303688, "learning_rate": 3.8218231213777486e-05, "loss": 0.0202, "step": 95510 }, { "epoch": 1.9087966108468886, "grad_norm": 0.08964091539382935, "learning_rate": 3.820818827395825e-05, "loss": 0.0272, "step": 95520 }, { "epoch": 1.90899644298789, "grad_norm": 0.11357882618904114, "learning_rate": 3.819814583786354e-05, "loss": 0.031, "step": 95530 }, { "epoch": 1.9091962751288918, "grad_norm": 0.14011995494365692, "learning_rate": 3.818810390592233e-05, "loss": 0.0232, "step": 95540 }, { "epoch": 1.9093961072698933, "grad_norm": 0.13489075005054474, "learning_rate": 3.8178062478563656e-05, "loss": 0.0287, "step": 95550 }, { "epoch": 1.9095959394108948, "grad_norm": 0.13684532046318054, "learning_rate": 3.8168021556216425e-05, "loss": 0.0212, "step": 95560 }, { "epoch": 1.9097957715518965, "grad_norm": 0.10682149231433868, "learning_rate": 3.815798113930959e-05, "loss": 0.0275, "step": 95570 }, { "epoch": 1.909995603692898, "grad_norm": 0.12519170343875885, "learning_rate": 3.8147941228272025e-05, "loss": 0.038, "step": 95580 }, { "epoch": 1.9101954358338995, "grad_norm": 0.1437968611717224, "learning_rate": 3.813790182353264e-05, "loss": 0.0277, "step": 95590 }, { "epoch": 1.9103952679749012, "grad_norm": 0.09070142358541489, "learning_rate": 3.8127862925520315e-05, "loss": 0.0294, "step": 95600 }, { "epoch": 1.9105951001159025, "grad_norm": 0.08606895059347153, "learning_rate": 3.811782453466386e-05, "loss": 0.0271, "step": 95610 }, { "epoch": 1.9107949322569042, "grad_norm": 0.14254328608512878, "learning_rate": 3.810778665139213e-05, "loss": 0.0285, "step": 95620 }, { "epoch": 1.9109947643979057, "grad_norm": 0.18280676007270813, "learning_rate": 3.809774927613388e-05, "loss": 0.0296, "step": 95630 }, { "epoch": 1.9111945965389072, "grad_norm": 0.1455090492963791, "learning_rate": 3.8087712409317936e-05, "loss": 0.0246, "step": 95640 }, { "epoch": 1.911394428679909, "grad_norm": 0.11374311149120331, "learning_rate": 3.8077676051373033e-05, "loss": 0.024, "step": 95650 }, { "epoch": 1.9115942608209104, "grad_norm": 0.10559520125389099, "learning_rate": 3.806764020272791e-05, "loss": 0.0241, "step": 95660 }, { "epoch": 1.911794092961912, "grad_norm": 0.11709544062614441, "learning_rate": 3.805760486381125e-05, "loss": 0.0229, "step": 95670 }, { "epoch": 1.9119939251029137, "grad_norm": 0.11195337027311325, "learning_rate": 3.8047570035051805e-05, "loss": 0.0274, "step": 95680 }, { "epoch": 1.9121937572439152, "grad_norm": 0.13173441588878632, "learning_rate": 3.8037535716878195e-05, "loss": 0.0252, "step": 95690 }, { "epoch": 1.9123935893849167, "grad_norm": 0.08357619494199753, "learning_rate": 3.8027501909719095e-05, "loss": 0.0264, "step": 95700 }, { "epoch": 1.9125934215259184, "grad_norm": 0.12864835560321808, "learning_rate": 3.80174686140031e-05, "loss": 0.0285, "step": 95710 }, { "epoch": 1.9127932536669197, "grad_norm": 0.1427733451128006, "learning_rate": 3.800743583015884e-05, "loss": 0.0326, "step": 95720 }, { "epoch": 1.9129930858079214, "grad_norm": 0.16178862750530243, "learning_rate": 3.79974035586149e-05, "loss": 0.0276, "step": 95730 }, { "epoch": 1.9131929179489229, "grad_norm": 0.14451788365840912, "learning_rate": 3.7987371799799824e-05, "loss": 0.0253, "step": 95740 }, { "epoch": 1.9133927500899244, "grad_norm": 0.09769003093242645, "learning_rate": 3.7977340554142164e-05, "loss": 0.0284, "step": 95750 }, { "epoch": 1.913592582230926, "grad_norm": 0.1312108337879181, "learning_rate": 3.79673098220704e-05, "loss": 0.0262, "step": 95760 }, { "epoch": 1.9137924143719276, "grad_norm": 0.13526923954486847, "learning_rate": 3.795727960401309e-05, "loss": 0.0267, "step": 95770 }, { "epoch": 1.913992246512929, "grad_norm": 0.1353251188993454, "learning_rate": 3.7947249900398654e-05, "loss": 0.03, "step": 95780 }, { "epoch": 1.9141920786539308, "grad_norm": 0.11244501918554306, "learning_rate": 3.793722071165558e-05, "loss": 0.0249, "step": 95790 }, { "epoch": 1.914391910794932, "grad_norm": 0.19352693855762482, "learning_rate": 3.792719203821225e-05, "loss": 0.0383, "step": 95800 }, { "epoch": 1.9145917429359338, "grad_norm": 0.11005434393882751, "learning_rate": 3.791716388049711e-05, "loss": 0.0306, "step": 95810 }, { "epoch": 1.9147915750769355, "grad_norm": 0.13553503155708313, "learning_rate": 3.790713623893855e-05, "loss": 0.0347, "step": 95820 }, { "epoch": 1.9149914072179368, "grad_norm": 0.1192372515797615, "learning_rate": 3.789710911396489e-05, "loss": 0.0311, "step": 95830 }, { "epoch": 1.9151912393589385, "grad_norm": 0.1812412440776825, "learning_rate": 3.788708250600451e-05, "loss": 0.0294, "step": 95840 }, { "epoch": 1.91539107149994, "grad_norm": 0.16311460733413696, "learning_rate": 3.7877056415485697e-05, "loss": 0.0292, "step": 95850 }, { "epoch": 1.9155909036409415, "grad_norm": 0.14241455495357513, "learning_rate": 3.786703084283677e-05, "loss": 0.0334, "step": 95860 }, { "epoch": 1.9157907357819433, "grad_norm": 0.1210467666387558, "learning_rate": 3.785700578848601e-05, "loss": 0.0333, "step": 95870 }, { "epoch": 1.9159905679229448, "grad_norm": 0.1591600924730301, "learning_rate": 3.784698125286164e-05, "loss": 0.0257, "step": 95880 }, { "epoch": 1.9161904000639463, "grad_norm": 0.15117624402046204, "learning_rate": 3.78369572363919e-05, "loss": 0.0245, "step": 95890 }, { "epoch": 1.916390232204948, "grad_norm": 0.1127575933933258, "learning_rate": 3.782693373950502e-05, "loss": 0.0328, "step": 95900 }, { "epoch": 1.9165900643459493, "grad_norm": 0.11303193867206573, "learning_rate": 3.781691076262916e-05, "loss": 0.0225, "step": 95910 }, { "epoch": 1.916789896486951, "grad_norm": 0.16896066069602966, "learning_rate": 3.78068883061925e-05, "loss": 0.0341, "step": 95920 }, { "epoch": 1.9169897286279525, "grad_norm": 0.1293124258518219, "learning_rate": 3.779686637062316e-05, "loss": 0.0304, "step": 95930 }, { "epoch": 1.917189560768954, "grad_norm": 0.08419033139944077, "learning_rate": 3.778684495634927e-05, "loss": 0.0245, "step": 95940 }, { "epoch": 1.9173893929099557, "grad_norm": 0.11389517039060593, "learning_rate": 3.7776824063798934e-05, "loss": 0.0255, "step": 95950 }, { "epoch": 1.9175892250509572, "grad_norm": 0.10001032799482346, "learning_rate": 3.776680369340021e-05, "loss": 0.023, "step": 95960 }, { "epoch": 1.9177890571919587, "grad_norm": 0.1309637874364853, "learning_rate": 3.775678384558118e-05, "loss": 0.0229, "step": 95970 }, { "epoch": 1.9179888893329604, "grad_norm": 0.09352574497461319, "learning_rate": 3.7746764520769826e-05, "loss": 0.0245, "step": 95980 }, { "epoch": 1.918188721473962, "grad_norm": 0.17488518357276917, "learning_rate": 3.773674571939419e-05, "loss": 0.0264, "step": 95990 }, { "epoch": 1.9183885536149634, "grad_norm": 0.15327076613903046, "learning_rate": 3.772672744188226e-05, "loss": 0.0293, "step": 96000 }, { "epoch": 1.9185883857559651, "grad_norm": 0.13394707441329956, "learning_rate": 3.7716709688661986e-05, "loss": 0.0334, "step": 96010 }, { "epoch": 1.9187882178969664, "grad_norm": 0.11714146286249161, "learning_rate": 3.770669246016131e-05, "loss": 0.0281, "step": 96020 }, { "epoch": 1.9189880500379681, "grad_norm": 0.1452764868736267, "learning_rate": 3.769667575680812e-05, "loss": 0.0347, "step": 96030 }, { "epoch": 1.9191878821789696, "grad_norm": 0.15922899544239044, "learning_rate": 3.768665957903037e-05, "loss": 0.023, "step": 96040 }, { "epoch": 1.9193877143199711, "grad_norm": 0.15642225742340088, "learning_rate": 3.76766439272559e-05, "loss": 0.0294, "step": 96050 }, { "epoch": 1.9195875464609728, "grad_norm": 0.11836940795183182, "learning_rate": 3.766662880191255e-05, "loss": 0.0294, "step": 96060 }, { "epoch": 1.9197873786019743, "grad_norm": 0.14641477167606354, "learning_rate": 3.765661420342816e-05, "loss": 0.0228, "step": 96070 }, { "epoch": 1.9199872107429758, "grad_norm": 0.11169911175966263, "learning_rate": 3.764660013223055e-05, "loss": 0.0291, "step": 96080 }, { "epoch": 1.9201870428839776, "grad_norm": 0.151312917470932, "learning_rate": 3.763658658874748e-05, "loss": 0.0323, "step": 96090 }, { "epoch": 1.920386875024979, "grad_norm": 0.1335982084274292, "learning_rate": 3.762657357340673e-05, "loss": 0.023, "step": 96100 }, { "epoch": 1.9205867071659806, "grad_norm": 0.12104154378175735, "learning_rate": 3.761656108663601e-05, "loss": 0.0227, "step": 96110 }, { "epoch": 1.9207865393069823, "grad_norm": 0.11411049216985703, "learning_rate": 3.7606549128863046e-05, "loss": 0.0274, "step": 96120 }, { "epoch": 1.9209863714479836, "grad_norm": 0.15190500020980835, "learning_rate": 3.759653770051555e-05, "loss": 0.0292, "step": 96130 }, { "epoch": 1.9211862035889853, "grad_norm": 0.09409811347723007, "learning_rate": 3.758652680202118e-05, "loss": 0.0211, "step": 96140 }, { "epoch": 1.9213860357299868, "grad_norm": 0.30779534578323364, "learning_rate": 3.7576516433807584e-05, "loss": 0.0233, "step": 96150 }, { "epoch": 1.9215858678709883, "grad_norm": 0.08915545046329498, "learning_rate": 3.756650659630235e-05, "loss": 0.0256, "step": 96160 }, { "epoch": 1.92178570001199, "grad_norm": 0.13059264421463013, "learning_rate": 3.755649728993315e-05, "loss": 0.0252, "step": 96170 }, { "epoch": 1.9219855321529915, "grad_norm": 0.10935889184474945, "learning_rate": 3.754648851512751e-05, "loss": 0.0345, "step": 96180 }, { "epoch": 1.922185364293993, "grad_norm": 0.09242703765630722, "learning_rate": 3.7536480272313016e-05, "loss": 0.0213, "step": 96190 }, { "epoch": 1.9223851964349947, "grad_norm": 0.15294277667999268, "learning_rate": 3.7526472561917175e-05, "loss": 0.0267, "step": 96200 }, { "epoch": 1.922585028575996, "grad_norm": 0.1438051015138626, "learning_rate": 3.75164653843675e-05, "loss": 0.021, "step": 96210 }, { "epoch": 1.9227848607169977, "grad_norm": 0.1491186022758484, "learning_rate": 3.7506458740091523e-05, "loss": 0.0297, "step": 96220 }, { "epoch": 1.9229846928579994, "grad_norm": 0.13959135115146637, "learning_rate": 3.749645262951665e-05, "loss": 0.0363, "step": 96230 }, { "epoch": 1.9231845249990007, "grad_norm": 0.16411103308200836, "learning_rate": 3.748644705307037e-05, "loss": 0.0261, "step": 96240 }, { "epoch": 1.9233843571400024, "grad_norm": 0.16361278295516968, "learning_rate": 3.747644201118007e-05, "loss": 0.0374, "step": 96250 }, { "epoch": 1.923584189281004, "grad_norm": 0.110403873026371, "learning_rate": 3.746643750427317e-05, "loss": 0.0245, "step": 96260 }, { "epoch": 1.9237840214220054, "grad_norm": 0.12949486076831818, "learning_rate": 3.7456433532777035e-05, "loss": 0.0266, "step": 96270 }, { "epoch": 1.9239838535630072, "grad_norm": 0.11266273260116577, "learning_rate": 3.744643009711901e-05, "loss": 0.0221, "step": 96280 }, { "epoch": 1.9241836857040087, "grad_norm": 0.10509410500526428, "learning_rate": 3.7436427197726445e-05, "loss": 0.0231, "step": 96290 }, { "epoch": 1.9243835178450102, "grad_norm": 0.12131920456886292, "learning_rate": 3.74264248350266e-05, "loss": 0.0331, "step": 96300 }, { "epoch": 1.9245833499860119, "grad_norm": 0.16298606991767883, "learning_rate": 3.74164230094468e-05, "loss": 0.0271, "step": 96310 }, { "epoch": 1.9247831821270132, "grad_norm": 0.11645690351724625, "learning_rate": 3.740642172141431e-05, "loss": 0.0308, "step": 96320 }, { "epoch": 1.9249830142680149, "grad_norm": 0.12128113210201263, "learning_rate": 3.739642097135633e-05, "loss": 0.0193, "step": 96330 }, { "epoch": 1.9251828464090164, "grad_norm": 0.13986097276210785, "learning_rate": 3.738642075970009e-05, "loss": 0.0345, "step": 96340 }, { "epoch": 1.9253826785500179, "grad_norm": 0.1366918534040451, "learning_rate": 3.73764210868728e-05, "loss": 0.0215, "step": 96350 }, { "epoch": 1.9255825106910196, "grad_norm": 0.11970203369855881, "learning_rate": 3.73664219533016e-05, "loss": 0.0295, "step": 96360 }, { "epoch": 1.925782342832021, "grad_norm": 0.1363745927810669, "learning_rate": 3.7356423359413654e-05, "loss": 0.0361, "step": 96370 }, { "epoch": 1.9259821749730226, "grad_norm": 0.11425454914569855, "learning_rate": 3.734642530563607e-05, "loss": 0.0271, "step": 96380 }, { "epoch": 1.9261820071140243, "grad_norm": 0.12368655949831009, "learning_rate": 3.733642779239594e-05, "loss": 0.0288, "step": 96390 }, { "epoch": 1.9263818392550258, "grad_norm": 0.10115589201450348, "learning_rate": 3.732643082012037e-05, "loss": 0.0287, "step": 96400 }, { "epoch": 1.9265816713960273, "grad_norm": 0.09019647538661957, "learning_rate": 3.731643438923638e-05, "loss": 0.0238, "step": 96410 }, { "epoch": 1.926781503537029, "grad_norm": 0.14092585444450378, "learning_rate": 3.730643850017103e-05, "loss": 0.0232, "step": 96420 }, { "epoch": 1.9269813356780303, "grad_norm": 0.13269206881523132, "learning_rate": 3.729644315335128e-05, "loss": 0.0381, "step": 96430 }, { "epoch": 1.927181167819032, "grad_norm": 0.11575797945261002, "learning_rate": 3.728644834920416e-05, "loss": 0.0208, "step": 96440 }, { "epoch": 1.9273809999600335, "grad_norm": 0.12987367808818817, "learning_rate": 3.7276454088156616e-05, "loss": 0.023, "step": 96450 }, { "epoch": 1.927580832101035, "grad_norm": 0.10314980149269104, "learning_rate": 3.726646037063557e-05, "loss": 0.0268, "step": 96460 }, { "epoch": 1.9277806642420368, "grad_norm": 0.14402435719966888, "learning_rate": 3.7256467197067956e-05, "loss": 0.0271, "step": 96470 }, { "epoch": 1.9279804963830383, "grad_norm": 0.10955261439085007, "learning_rate": 3.724647456788063e-05, "loss": 0.0259, "step": 96480 }, { "epoch": 1.9281803285240398, "grad_norm": 0.13637475669384003, "learning_rate": 3.723648248350049e-05, "loss": 0.0294, "step": 96490 }, { "epoch": 1.9283801606650415, "grad_norm": 0.10517371445894241, "learning_rate": 3.722649094435439e-05, "loss": 0.0216, "step": 96500 }, { "epoch": 1.928579992806043, "grad_norm": 0.09755664318799973, "learning_rate": 3.721649995086911e-05, "loss": 0.0332, "step": 96510 }, { "epoch": 1.9287798249470445, "grad_norm": 0.16392190754413605, "learning_rate": 3.7206509503471466e-05, "loss": 0.0259, "step": 96520 }, { "epoch": 1.9289796570880462, "grad_norm": 0.16699813306331635, "learning_rate": 3.719651960258825e-05, "loss": 0.0303, "step": 96530 }, { "epoch": 1.9291794892290475, "grad_norm": 0.11992615461349487, "learning_rate": 3.718653024864619e-05, "loss": 0.028, "step": 96540 }, { "epoch": 1.9293793213700492, "grad_norm": 0.09382548183202744, "learning_rate": 3.717654144207202e-05, "loss": 0.0243, "step": 96550 }, { "epoch": 1.9295791535110507, "grad_norm": 0.12200169265270233, "learning_rate": 3.716655318329244e-05, "loss": 0.023, "step": 96560 }, { "epoch": 1.9297789856520522, "grad_norm": 0.1866215020418167, "learning_rate": 3.7156565472734116e-05, "loss": 0.0231, "step": 96570 }, { "epoch": 1.929978817793054, "grad_norm": 0.13570941984653473, "learning_rate": 3.7146578310823735e-05, "loss": 0.0256, "step": 96580 }, { "epoch": 1.9301786499340554, "grad_norm": 0.09878307580947876, "learning_rate": 3.7136591697987934e-05, "loss": 0.026, "step": 96590 }, { "epoch": 1.930378482075057, "grad_norm": 0.13565313816070557, "learning_rate": 3.712660563465328e-05, "loss": 0.0251, "step": 96600 }, { "epoch": 1.9305783142160586, "grad_norm": 0.18317490816116333, "learning_rate": 3.711662012124638e-05, "loss": 0.0319, "step": 96610 }, { "epoch": 1.93077814635706, "grad_norm": 0.10438954830169678, "learning_rate": 3.710663515819383e-05, "loss": 0.0243, "step": 96620 }, { "epoch": 1.9309779784980616, "grad_norm": 0.13044708967208862, "learning_rate": 3.709665074592211e-05, "loss": 0.0298, "step": 96630 }, { "epoch": 1.9311778106390634, "grad_norm": 0.12620475888252258, "learning_rate": 3.708666688485779e-05, "loss": 0.0254, "step": 96640 }, { "epoch": 1.9313776427800646, "grad_norm": 0.11602940410375595, "learning_rate": 3.707668357542733e-05, "loss": 0.0199, "step": 96650 }, { "epoch": 1.9315774749210664, "grad_norm": 0.14045079052448273, "learning_rate": 3.7066700818057185e-05, "loss": 0.0381, "step": 96660 }, { "epoch": 1.9317773070620678, "grad_norm": 0.14704753458499908, "learning_rate": 3.7056718613173855e-05, "loss": 0.033, "step": 96670 }, { "epoch": 1.9319771392030693, "grad_norm": 0.14217428863048553, "learning_rate": 3.704673696120371e-05, "loss": 0.0303, "step": 96680 }, { "epoch": 1.932176971344071, "grad_norm": 0.11867411434650421, "learning_rate": 3.703675586257318e-05, "loss": 0.0308, "step": 96690 }, { "epoch": 1.9323768034850726, "grad_norm": 0.09046164155006409, "learning_rate": 3.702677531770861e-05, "loss": 0.0223, "step": 96700 }, { "epoch": 1.932576635626074, "grad_norm": 0.1392778605222702, "learning_rate": 3.7016795327036366e-05, "loss": 0.0325, "step": 96710 }, { "epoch": 1.9327764677670758, "grad_norm": 0.11596096307039261, "learning_rate": 3.700681589098279e-05, "loss": 0.0332, "step": 96720 }, { "epoch": 1.932976299908077, "grad_norm": 0.09895123541355133, "learning_rate": 3.699683700997416e-05, "loss": 0.021, "step": 96730 }, { "epoch": 1.9331761320490788, "grad_norm": 0.16154131293296814, "learning_rate": 3.698685868443677e-05, "loss": 0.0226, "step": 96740 }, { "epoch": 1.9333759641900803, "grad_norm": 0.16849379241466522, "learning_rate": 3.697688091479685e-05, "loss": 0.0267, "step": 96750 }, { "epoch": 1.9335757963310818, "grad_norm": 0.11260449886322021, "learning_rate": 3.696690370148067e-05, "loss": 0.0248, "step": 96760 }, { "epoch": 1.9337756284720835, "grad_norm": 0.1264084130525589, "learning_rate": 3.6956927044914436e-05, "loss": 0.0352, "step": 96770 }, { "epoch": 1.933975460613085, "grad_norm": 0.13741710782051086, "learning_rate": 3.694695094552429e-05, "loss": 0.0304, "step": 96780 }, { "epoch": 1.9341752927540865, "grad_norm": 0.09153182059526443, "learning_rate": 3.693697540373643e-05, "loss": 0.0251, "step": 96790 }, { "epoch": 1.9343751248950882, "grad_norm": 0.1203811839222908, "learning_rate": 3.6927000419976995e-05, "loss": 0.024, "step": 96800 }, { "epoch": 1.9345749570360897, "grad_norm": 0.10824360698461533, "learning_rate": 3.691702599467207e-05, "loss": 0.0258, "step": 96810 }, { "epoch": 1.9347747891770912, "grad_norm": 0.1304241269826889, "learning_rate": 3.690705212824778e-05, "loss": 0.0254, "step": 96820 }, { "epoch": 1.934974621318093, "grad_norm": 0.19805969297885895, "learning_rate": 3.689707882113016e-05, "loss": 0.0293, "step": 96830 }, { "epoch": 1.9351744534590942, "grad_norm": 0.12225408852100372, "learning_rate": 3.688710607374525e-05, "loss": 0.0357, "step": 96840 }, { "epoch": 1.935374285600096, "grad_norm": 0.12398167699575424, "learning_rate": 3.687713388651909e-05, "loss": 0.0235, "step": 96850 }, { "epoch": 1.9355741177410974, "grad_norm": 0.08539851009845734, "learning_rate": 3.6867162259877665e-05, "loss": 0.0293, "step": 96860 }, { "epoch": 1.935773949882099, "grad_norm": 0.11107071489095688, "learning_rate": 3.685719119424695e-05, "loss": 0.0309, "step": 96870 }, { "epoch": 1.9359737820231007, "grad_norm": 0.15367738902568817, "learning_rate": 3.6847220690052856e-05, "loss": 0.0276, "step": 96880 }, { "epoch": 1.9361736141641022, "grad_norm": 0.14385254681110382, "learning_rate": 3.6837250747721353e-05, "loss": 0.0282, "step": 96890 }, { "epoch": 1.9363734463051037, "grad_norm": 0.13877983391284943, "learning_rate": 3.682728136767832e-05, "loss": 0.0312, "step": 96900 }, { "epoch": 1.9365732784461054, "grad_norm": 0.12790624797344208, "learning_rate": 3.681731255034962e-05, "loss": 0.0258, "step": 96910 }, { "epoch": 1.9367731105871069, "grad_norm": 0.08610305190086365, "learning_rate": 3.6807344296161116e-05, "loss": 0.0204, "step": 96920 }, { "epoch": 1.9369729427281084, "grad_norm": 0.13448765873908997, "learning_rate": 3.6797376605538605e-05, "loss": 0.0339, "step": 96930 }, { "epoch": 1.93717277486911, "grad_norm": 0.13119564950466156, "learning_rate": 3.678740947890794e-05, "loss": 0.0217, "step": 96940 }, { "epoch": 1.9373726070101114, "grad_norm": 0.13454777002334595, "learning_rate": 3.677744291669487e-05, "loss": 0.024, "step": 96950 }, { "epoch": 1.937572439151113, "grad_norm": 0.13681067526340485, "learning_rate": 3.676747691932513e-05, "loss": 0.0214, "step": 96960 }, { "epoch": 1.9377722712921146, "grad_norm": 0.124944768846035, "learning_rate": 3.675751148722446e-05, "loss": 0.0252, "step": 96970 }, { "epoch": 1.937972103433116, "grad_norm": 0.10939224064350128, "learning_rate": 3.674754662081859e-05, "loss": 0.0279, "step": 96980 }, { "epoch": 1.9381719355741178, "grad_norm": 0.11828868836164474, "learning_rate": 3.673758232053319e-05, "loss": 0.0318, "step": 96990 }, { "epoch": 1.9383717677151193, "grad_norm": 0.1848781257867813, "learning_rate": 3.672761858679389e-05, "loss": 0.0259, "step": 97000 }, { "epoch": 1.9385715998561208, "grad_norm": 0.1071552261710167, "learning_rate": 3.6717655420026345e-05, "loss": 0.0284, "step": 97010 }, { "epoch": 1.9387714319971225, "grad_norm": 0.12591855227947235, "learning_rate": 3.670769282065617e-05, "loss": 0.027, "step": 97020 }, { "epoch": 1.9389712641381238, "grad_norm": 0.14520445466041565, "learning_rate": 3.669773078910894e-05, "loss": 0.0267, "step": 97030 }, { "epoch": 1.9391710962791255, "grad_norm": 0.14182773232460022, "learning_rate": 3.668776932581022e-05, "loss": 0.0242, "step": 97040 }, { "epoch": 1.939370928420127, "grad_norm": 0.07955389469861984, "learning_rate": 3.667780843118552e-05, "loss": 0.0241, "step": 97050 }, { "epoch": 1.9395707605611285, "grad_norm": 0.1430085450410843, "learning_rate": 3.666784810566036e-05, "loss": 0.0255, "step": 97060 }, { "epoch": 1.9397705927021303, "grad_norm": 0.16459324955940247, "learning_rate": 3.665788834966027e-05, "loss": 0.0325, "step": 97070 }, { "epoch": 1.9399704248431318, "grad_norm": 0.12082737684249878, "learning_rate": 3.664792916361066e-05, "loss": 0.0241, "step": 97080 }, { "epoch": 1.9401702569841333, "grad_norm": 0.10744684189558029, "learning_rate": 3.6637970547937e-05, "loss": 0.0281, "step": 97090 }, { "epoch": 1.940370089125135, "grad_norm": 0.10560619831085205, "learning_rate": 3.662801250306467e-05, "loss": 0.0285, "step": 97100 }, { "epoch": 1.9405699212661365, "grad_norm": 0.24274717271327972, "learning_rate": 3.66180550294191e-05, "loss": 0.0309, "step": 97110 }, { "epoch": 1.940769753407138, "grad_norm": 0.13157832622528076, "learning_rate": 3.660809812742563e-05, "loss": 0.0217, "step": 97120 }, { "epoch": 1.9409695855481397, "grad_norm": 0.20057834684848785, "learning_rate": 3.659814179750961e-05, "loss": 0.0289, "step": 97130 }, { "epoch": 1.941169417689141, "grad_norm": 0.0994143933057785, "learning_rate": 3.658818604009635e-05, "loss": 0.0277, "step": 97140 }, { "epoch": 1.9413692498301427, "grad_norm": 0.13600978255271912, "learning_rate": 3.657823085561113e-05, "loss": 0.0273, "step": 97150 }, { "epoch": 1.9415690819711442, "grad_norm": 0.11778335273265839, "learning_rate": 3.656827624447925e-05, "loss": 0.034, "step": 97160 }, { "epoch": 1.9417689141121457, "grad_norm": 0.149805948138237, "learning_rate": 3.655832220712593e-05, "loss": 0.0304, "step": 97170 }, { "epoch": 1.9419687462531474, "grad_norm": 0.12342716753482819, "learning_rate": 3.654836874397639e-05, "loss": 0.0275, "step": 97180 }, { "epoch": 1.942168578394149, "grad_norm": 0.11808838695287704, "learning_rate": 3.653841585545581e-05, "loss": 0.0342, "step": 97190 }, { "epoch": 1.9423684105351504, "grad_norm": 0.10843438655138016, "learning_rate": 3.65284635419894e-05, "loss": 0.0325, "step": 97200 }, { "epoch": 1.9425682426761521, "grad_norm": 0.13444501161575317, "learning_rate": 3.651851180400227e-05, "loss": 0.0341, "step": 97210 }, { "epoch": 1.9427680748171536, "grad_norm": 0.11581810563802719, "learning_rate": 3.650856064191956e-05, "loss": 0.0327, "step": 97220 }, { "epoch": 1.9429679069581551, "grad_norm": 0.1423576921224594, "learning_rate": 3.649861005616634e-05, "loss": 0.0278, "step": 97230 }, { "epoch": 1.9431677390991569, "grad_norm": 0.15909862518310547, "learning_rate": 3.6488660047167686e-05, "loss": 0.0272, "step": 97240 }, { "epoch": 1.9433675712401581, "grad_norm": 0.19612792134284973, "learning_rate": 3.6478710615348675e-05, "loss": 0.0272, "step": 97250 }, { "epoch": 1.9435674033811599, "grad_norm": 0.14254403114318848, "learning_rate": 3.64687617611343e-05, "loss": 0.0332, "step": 97260 }, { "epoch": 1.9437672355221614, "grad_norm": 0.1424279361963272, "learning_rate": 3.645881348494957e-05, "loss": 0.0294, "step": 97270 }, { "epoch": 1.9439670676631629, "grad_norm": 0.10434727370738983, "learning_rate": 3.644886578721944e-05, "loss": 0.0268, "step": 97280 }, { "epoch": 1.9441668998041646, "grad_norm": 0.10623869299888611, "learning_rate": 3.6438918668368874e-05, "loss": 0.0265, "step": 97290 }, { "epoch": 1.944366731945166, "grad_norm": 0.1398071050643921, "learning_rate": 3.6428972128822805e-05, "loss": 0.0208, "step": 97300 }, { "epoch": 1.9445665640861676, "grad_norm": 0.1552889049053192, "learning_rate": 3.64190261690061e-05, "loss": 0.0289, "step": 97310 }, { "epoch": 1.9447663962271693, "grad_norm": 0.12578605115413666, "learning_rate": 3.640908078934365e-05, "loss": 0.0262, "step": 97320 }, { "epoch": 1.9449662283681708, "grad_norm": 0.11156260967254639, "learning_rate": 3.639913599026028e-05, "loss": 0.0371, "step": 97330 }, { "epoch": 1.9451660605091723, "grad_norm": 0.12088123708963394, "learning_rate": 3.638919177218087e-05, "loss": 0.0203, "step": 97340 }, { "epoch": 1.945365892650174, "grad_norm": 0.1415235698223114, "learning_rate": 3.637924813553017e-05, "loss": 0.0278, "step": 97350 }, { "epoch": 1.9455657247911753, "grad_norm": 0.10444004088640213, "learning_rate": 3.636930508073297e-05, "loss": 0.035, "step": 97360 }, { "epoch": 1.945765556932177, "grad_norm": 0.10465541481971741, "learning_rate": 3.635936260821401e-05, "loss": 0.0283, "step": 97370 }, { "epoch": 1.9459653890731785, "grad_norm": 0.14458051323890686, "learning_rate": 3.634942071839802e-05, "loss": 0.026, "step": 97380 }, { "epoch": 1.94616522121418, "grad_norm": 0.13215702772140503, "learning_rate": 3.633947941170972e-05, "loss": 0.0307, "step": 97390 }, { "epoch": 1.9463650533551817, "grad_norm": 0.1146499291062355, "learning_rate": 3.6329538688573746e-05, "loss": 0.0303, "step": 97400 }, { "epoch": 1.9465648854961832, "grad_norm": 0.13711850345134735, "learning_rate": 3.631959854941478e-05, "loss": 0.0317, "step": 97410 }, { "epoch": 1.9467647176371847, "grad_norm": 0.11978454887866974, "learning_rate": 3.63096589946574e-05, "loss": 0.0294, "step": 97420 }, { "epoch": 1.9469645497781864, "grad_norm": 0.132737398147583, "learning_rate": 3.629972002472626e-05, "loss": 0.0309, "step": 97430 }, { "epoch": 1.9471643819191877, "grad_norm": 0.10599704086780548, "learning_rate": 3.6289781640045926e-05, "loss": 0.0253, "step": 97440 }, { "epoch": 1.9473642140601894, "grad_norm": 0.12003076821565628, "learning_rate": 3.6279843841040916e-05, "loss": 0.0254, "step": 97450 }, { "epoch": 1.947564046201191, "grad_norm": 0.12969014048576355, "learning_rate": 3.626990662813576e-05, "loss": 0.0312, "step": 97460 }, { "epoch": 1.9477638783421924, "grad_norm": 0.11061502248048782, "learning_rate": 3.6259970001755e-05, "loss": 0.0232, "step": 97470 }, { "epoch": 1.9479637104831942, "grad_norm": 0.16284285485744476, "learning_rate": 3.625003396232306e-05, "loss": 0.0284, "step": 97480 }, { "epoch": 1.9481635426241957, "grad_norm": 0.11621911823749542, "learning_rate": 3.624009851026443e-05, "loss": 0.0216, "step": 97490 }, { "epoch": 1.9483633747651972, "grad_norm": 0.11735610663890839, "learning_rate": 3.62301636460035e-05, "loss": 0.0275, "step": 97500 }, { "epoch": 1.9485632069061989, "grad_norm": 0.14165426790714264, "learning_rate": 3.622022936996467e-05, "loss": 0.0239, "step": 97510 }, { "epoch": 1.9487630390472004, "grad_norm": 0.13524290919303894, "learning_rate": 3.621029568257235e-05, "loss": 0.0288, "step": 97520 }, { "epoch": 1.9489628711882019, "grad_norm": 0.12510570883750916, "learning_rate": 3.6200362584250844e-05, "loss": 0.0267, "step": 97530 }, { "epoch": 1.9491627033292036, "grad_norm": 0.12056656926870346, "learning_rate": 3.619043007542452e-05, "loss": 0.0216, "step": 97540 }, { "epoch": 1.9493625354702049, "grad_norm": 0.11976537853479385, "learning_rate": 3.618049815651763e-05, "loss": 0.029, "step": 97550 }, { "epoch": 1.9495623676112066, "grad_norm": 0.12967386841773987, "learning_rate": 3.617056682795449e-05, "loss": 0.0306, "step": 97560 }, { "epoch": 1.949762199752208, "grad_norm": 0.09173430502414703, "learning_rate": 3.616063609015933e-05, "loss": 0.033, "step": 97570 }, { "epoch": 1.9499620318932096, "grad_norm": 0.1143704205751419, "learning_rate": 3.615070594355637e-05, "loss": 0.0227, "step": 97580 }, { "epoch": 1.9501618640342113, "grad_norm": 0.1101384237408638, "learning_rate": 3.614077638856981e-05, "loss": 0.0281, "step": 97590 }, { "epoch": 1.9503616961752128, "grad_norm": 0.11076945811510086, "learning_rate": 3.6130847425623815e-05, "loss": 0.0354, "step": 97600 }, { "epoch": 1.9505615283162143, "grad_norm": 0.14532068371772766, "learning_rate": 3.6120919055142536e-05, "loss": 0.0444, "step": 97610 }, { "epoch": 1.950761360457216, "grad_norm": 0.11832185834646225, "learning_rate": 3.611099127755012e-05, "loss": 0.0281, "step": 97620 }, { "epoch": 1.9509611925982175, "grad_norm": 0.11584346741437912, "learning_rate": 3.610106409327063e-05, "loss": 0.0285, "step": 97630 }, { "epoch": 1.951161024739219, "grad_norm": 0.12413438409566879, "learning_rate": 3.6091137502728134e-05, "loss": 0.0275, "step": 97640 }, { "epoch": 1.9513608568802208, "grad_norm": 0.1961905062198639, "learning_rate": 3.6081211506346714e-05, "loss": 0.0289, "step": 97650 }, { "epoch": 1.951560689021222, "grad_norm": 0.11305877566337585, "learning_rate": 3.607128610455036e-05, "loss": 0.021, "step": 97660 }, { "epoch": 1.9517605211622238, "grad_norm": 0.10900717228651047, "learning_rate": 3.606136129776308e-05, "loss": 0.0279, "step": 97670 }, { "epoch": 1.9519603533032253, "grad_norm": 0.11165355145931244, "learning_rate": 3.605143708640883e-05, "loss": 0.0274, "step": 97680 }, { "epoch": 1.9521601854442268, "grad_norm": 0.12152333557605743, "learning_rate": 3.604151347091156e-05, "loss": 0.0271, "step": 97690 }, { "epoch": 1.9523600175852285, "grad_norm": 0.15638147294521332, "learning_rate": 3.603159045169521e-05, "loss": 0.0284, "step": 97700 }, { "epoch": 1.95255984972623, "grad_norm": 0.10366932302713394, "learning_rate": 3.602166802918364e-05, "loss": 0.0297, "step": 97710 }, { "epoch": 1.9527596818672315, "grad_norm": 0.115837462246418, "learning_rate": 3.6011746203800754e-05, "loss": 0.0331, "step": 97720 }, { "epoch": 1.9529595140082332, "grad_norm": 0.11377312242984772, "learning_rate": 3.600182497597033e-05, "loss": 0.0303, "step": 97730 }, { "epoch": 1.9531593461492345, "grad_norm": 0.10743190348148346, "learning_rate": 3.5991904346116264e-05, "loss": 0.0333, "step": 97740 }, { "epoch": 1.9533591782902362, "grad_norm": 0.14366869628429413, "learning_rate": 3.59819843146623e-05, "loss": 0.0267, "step": 97750 }, { "epoch": 1.953559010431238, "grad_norm": 0.10792543739080429, "learning_rate": 3.597206488203223e-05, "loss": 0.0282, "step": 97760 }, { "epoch": 1.9537588425722392, "grad_norm": 0.11805947124958038, "learning_rate": 3.5962146048649756e-05, "loss": 0.0391, "step": 97770 }, { "epoch": 1.953958674713241, "grad_norm": 0.14277291297912598, "learning_rate": 3.595222781493861e-05, "loss": 0.0346, "step": 97780 }, { "epoch": 1.9541585068542424, "grad_norm": 0.09788650274276733, "learning_rate": 3.594231018132251e-05, "loss": 0.0301, "step": 97790 }, { "epoch": 1.954358338995244, "grad_norm": 0.1335846185684204, "learning_rate": 3.593239314822508e-05, "loss": 0.0315, "step": 97800 }, { "epoch": 1.9545581711362456, "grad_norm": 0.14258606731891632, "learning_rate": 3.5922476716069986e-05, "loss": 0.0237, "step": 97810 }, { "epoch": 1.9547580032772471, "grad_norm": 0.15629729628562927, "learning_rate": 3.59125608852808e-05, "loss": 0.0221, "step": 97820 }, { "epoch": 1.9549578354182486, "grad_norm": 0.14216165244579315, "learning_rate": 3.590264565628114e-05, "loss": 0.025, "step": 97830 }, { "epoch": 1.9551576675592504, "grad_norm": 0.12517228722572327, "learning_rate": 3.5892731029494576e-05, "loss": 0.0221, "step": 97840 }, { "epoch": 1.9553574997002516, "grad_norm": 0.1126752495765686, "learning_rate": 3.5882817005344617e-05, "loss": 0.0226, "step": 97850 }, { "epoch": 1.9555573318412534, "grad_norm": 0.12531165778636932, "learning_rate": 3.5872903584254794e-05, "loss": 0.022, "step": 97860 }, { "epoch": 1.9557571639822549, "grad_norm": 0.15164116024971008, "learning_rate": 3.586299076664855e-05, "loss": 0.0311, "step": 97870 }, { "epoch": 1.9559569961232564, "grad_norm": 0.12724564969539642, "learning_rate": 3.585307855294938e-05, "loss": 0.0223, "step": 97880 }, { "epoch": 1.956156828264258, "grad_norm": 0.12021996080875397, "learning_rate": 3.584316694358072e-05, "loss": 0.0395, "step": 97890 }, { "epoch": 1.9563566604052596, "grad_norm": 0.14021585881710052, "learning_rate": 3.583325593896595e-05, "loss": 0.0257, "step": 97900 }, { "epoch": 1.956556492546261, "grad_norm": 0.16556228697299957, "learning_rate": 3.5823345539528454e-05, "loss": 0.0288, "step": 97910 }, { "epoch": 1.9567563246872628, "grad_norm": 0.1414152830839157, "learning_rate": 3.581343574569161e-05, "loss": 0.0262, "step": 97920 }, { "epoch": 1.9569561568282643, "grad_norm": 0.12396889925003052, "learning_rate": 3.5803526557878716e-05, "loss": 0.0256, "step": 97930 }, { "epoch": 1.9571559889692658, "grad_norm": 0.11768876761198044, "learning_rate": 3.57936179765131e-05, "loss": 0.0183, "step": 97940 }, { "epoch": 1.9573558211102675, "grad_norm": 0.21025437116622925, "learning_rate": 3.578371000201802e-05, "loss": 0.0264, "step": 97950 }, { "epoch": 1.9575556532512688, "grad_norm": 0.09868081659078598, "learning_rate": 3.577380263481672e-05, "loss": 0.0275, "step": 97960 }, { "epoch": 1.9577554853922705, "grad_norm": 0.15830272436141968, "learning_rate": 3.576389587533246e-05, "loss": 0.0254, "step": 97970 }, { "epoch": 1.957955317533272, "grad_norm": 0.09948789328336716, "learning_rate": 3.57539897239884e-05, "loss": 0.0334, "step": 97980 }, { "epoch": 1.9581551496742735, "grad_norm": 0.14141182601451874, "learning_rate": 3.574408418120775e-05, "loss": 0.0324, "step": 97990 }, { "epoch": 1.9583549818152752, "grad_norm": 0.11919591575860977, "learning_rate": 3.5734179247413604e-05, "loss": 0.0192, "step": 98000 }, { "epoch": 1.9585548139562767, "grad_norm": 0.11781527101993561, "learning_rate": 3.5724274923029124e-05, "loss": 0.024, "step": 98010 }, { "epoch": 1.9587546460972782, "grad_norm": 0.15996938943862915, "learning_rate": 3.5714371208477406e-05, "loss": 0.0335, "step": 98020 }, { "epoch": 1.95895447823828, "grad_norm": 0.12280932813882828, "learning_rate": 3.570446810418149e-05, "loss": 0.0258, "step": 98030 }, { "epoch": 1.9591543103792814, "grad_norm": 0.13565030694007874, "learning_rate": 3.5694565610564455e-05, "loss": 0.0236, "step": 98040 }, { "epoch": 1.959354142520283, "grad_norm": 0.15951576828956604, "learning_rate": 3.568466372804927e-05, "loss": 0.0291, "step": 98050 }, { "epoch": 1.9595539746612847, "grad_norm": 0.09907381981611252, "learning_rate": 3.567476245705896e-05, "loss": 0.026, "step": 98060 }, { "epoch": 1.959753806802286, "grad_norm": 0.14522454142570496, "learning_rate": 3.5664861798016504e-05, "loss": 0.0214, "step": 98070 }, { "epoch": 1.9599536389432877, "grad_norm": 0.093967504799366, "learning_rate": 3.5654961751344804e-05, "loss": 0.028, "step": 98080 }, { "epoch": 1.9601534710842892, "grad_norm": 0.10861098766326904, "learning_rate": 3.564506231746677e-05, "loss": 0.0223, "step": 98090 }, { "epoch": 1.9603533032252907, "grad_norm": 0.10919162631034851, "learning_rate": 3.563516349680532e-05, "loss": 0.0369, "step": 98100 }, { "epoch": 1.9605531353662924, "grad_norm": 0.11049294471740723, "learning_rate": 3.56252652897833e-05, "loss": 0.0278, "step": 98110 }, { "epoch": 1.9607529675072939, "grad_norm": 0.12831464409828186, "learning_rate": 3.5615367696823534e-05, "loss": 0.0223, "step": 98120 }, { "epoch": 1.9609527996482954, "grad_norm": 0.09374307841062546, "learning_rate": 3.560547071834883e-05, "loss": 0.0249, "step": 98130 }, { "epoch": 1.961152631789297, "grad_norm": 0.11245213449001312, "learning_rate": 3.559557435478196e-05, "loss": 0.0246, "step": 98140 }, { "epoch": 1.9613524639302984, "grad_norm": 0.12923674285411835, "learning_rate": 3.558567860654571e-05, "loss": 0.028, "step": 98150 }, { "epoch": 1.9615522960713, "grad_norm": 0.12310115247964859, "learning_rate": 3.55757834740628e-05, "loss": 0.0274, "step": 98160 }, { "epoch": 1.9617521282123018, "grad_norm": 0.16263142228126526, "learning_rate": 3.5565888957755907e-05, "loss": 0.0273, "step": 98170 }, { "epoch": 1.961951960353303, "grad_norm": 0.09885308891534805, "learning_rate": 3.5555995058047706e-05, "loss": 0.031, "step": 98180 }, { "epoch": 1.9621517924943048, "grad_norm": 0.15113727748394012, "learning_rate": 3.554610177536089e-05, "loss": 0.0289, "step": 98190 }, { "epoch": 1.9623516246353063, "grad_norm": 0.09861332178115845, "learning_rate": 3.553620911011805e-05, "loss": 0.0209, "step": 98200 }, { "epoch": 1.9625514567763078, "grad_norm": 0.16060411930084229, "learning_rate": 3.552631706274178e-05, "loss": 0.0245, "step": 98210 }, { "epoch": 1.9627512889173095, "grad_norm": 0.14752836525440216, "learning_rate": 3.551642563365464e-05, "loss": 0.0244, "step": 98220 }, { "epoch": 1.962951121058311, "grad_norm": 0.13736435770988464, "learning_rate": 3.5506534823279205e-05, "loss": 0.0295, "step": 98230 }, { "epoch": 1.9631509531993125, "grad_norm": 0.1293637454509735, "learning_rate": 3.549664463203799e-05, "loss": 0.0323, "step": 98240 }, { "epoch": 1.9633507853403143, "grad_norm": 0.08732011169195175, "learning_rate": 3.548675506035345e-05, "loss": 0.0241, "step": 98250 }, { "epoch": 1.9635506174813155, "grad_norm": 0.11096508055925369, "learning_rate": 3.54768661086481e-05, "loss": 0.0258, "step": 98260 }, { "epoch": 1.9637504496223173, "grad_norm": 0.1461714804172516, "learning_rate": 3.546697777734431e-05, "loss": 0.0379, "step": 98270 }, { "epoch": 1.9639502817633188, "grad_norm": 0.11993057280778885, "learning_rate": 3.545709006686456e-05, "loss": 0.0225, "step": 98280 }, { "epoch": 1.9641501139043203, "grad_norm": 0.08262763172388077, "learning_rate": 3.5447202977631205e-05, "loss": 0.0252, "step": 98290 }, { "epoch": 1.964349946045322, "grad_norm": 0.12317150831222534, "learning_rate": 3.5437316510066596e-05, "loss": 0.0279, "step": 98300 }, { "epoch": 1.9645497781863235, "grad_norm": 0.1438729465007782, "learning_rate": 3.542743066459306e-05, "loss": 0.0264, "step": 98310 }, { "epoch": 1.964749610327325, "grad_norm": 0.11238399893045425, "learning_rate": 3.541754544163295e-05, "loss": 0.0261, "step": 98320 }, { "epoch": 1.9649494424683267, "grad_norm": 0.16119670867919922, "learning_rate": 3.540766084160848e-05, "loss": 0.0303, "step": 98330 }, { "epoch": 1.9651492746093282, "grad_norm": 0.12760154902935028, "learning_rate": 3.539777686494196e-05, "loss": 0.0259, "step": 98340 }, { "epoch": 1.9653491067503297, "grad_norm": 0.14349591732025146, "learning_rate": 3.538789351205556e-05, "loss": 0.0273, "step": 98350 }, { "epoch": 1.9655489388913314, "grad_norm": 0.15400712192058563, "learning_rate": 3.537801078337151e-05, "loss": 0.0227, "step": 98360 }, { "epoch": 1.9657487710323327, "grad_norm": 0.09523427486419678, "learning_rate": 3.5368128679311995e-05, "loss": 0.0284, "step": 98370 }, { "epoch": 1.9659486031733344, "grad_norm": 0.11473383009433746, "learning_rate": 3.5358247200299126e-05, "loss": 0.0268, "step": 98380 }, { "epoch": 1.966148435314336, "grad_norm": 0.10859386622905731, "learning_rate": 3.534836634675506e-05, "loss": 0.029, "step": 98390 }, { "epoch": 1.9663482674553374, "grad_norm": 0.12904082238674164, "learning_rate": 3.5338486119101834e-05, "loss": 0.0287, "step": 98400 }, { "epoch": 1.9665480995963391, "grad_norm": 0.1385257989168167, "learning_rate": 3.532860651776156e-05, "loss": 0.0217, "step": 98410 }, { "epoch": 1.9667479317373406, "grad_norm": 0.12616552412509918, "learning_rate": 3.5318727543156284e-05, "loss": 0.0283, "step": 98420 }, { "epoch": 1.9669477638783421, "grad_norm": 0.13128049671649933, "learning_rate": 3.530884919570798e-05, "loss": 0.0267, "step": 98430 }, { "epoch": 1.9671475960193439, "grad_norm": 0.09773197025060654, "learning_rate": 3.529897147583866e-05, "loss": 0.0306, "step": 98440 }, { "epoch": 1.9673474281603454, "grad_norm": 0.12238427996635437, "learning_rate": 3.528909438397026e-05, "loss": 0.0235, "step": 98450 }, { "epoch": 1.9675472603013469, "grad_norm": 0.11744855344295502, "learning_rate": 3.527921792052472e-05, "loss": 0.0195, "step": 98460 }, { "epoch": 1.9677470924423486, "grad_norm": 0.1168479397892952, "learning_rate": 3.5269342085923964e-05, "loss": 0.0341, "step": 98470 }, { "epoch": 1.9679469245833499, "grad_norm": 0.15959316492080688, "learning_rate": 3.525946688058984e-05, "loss": 0.0304, "step": 98480 }, { "epoch": 1.9681467567243516, "grad_norm": 0.16652131080627441, "learning_rate": 3.52495923049442e-05, "loss": 0.0319, "step": 98490 }, { "epoch": 1.968346588865353, "grad_norm": 0.0845847874879837, "learning_rate": 3.5239718359408904e-05, "loss": 0.0238, "step": 98500 }, { "epoch": 1.9685464210063546, "grad_norm": 0.11758885532617569, "learning_rate": 3.522984504440572e-05, "loss": 0.0248, "step": 98510 }, { "epoch": 1.9687462531473563, "grad_norm": 0.1144808679819107, "learning_rate": 3.521997236035643e-05, "loss": 0.027, "step": 98520 }, { "epoch": 1.9689460852883578, "grad_norm": 0.12302415817975998, "learning_rate": 3.521010030768275e-05, "loss": 0.0212, "step": 98530 }, { "epoch": 1.9691459174293593, "grad_norm": 0.1493205577135086, "learning_rate": 3.52002288868064e-05, "loss": 0.029, "step": 98540 }, { "epoch": 1.969345749570361, "grad_norm": 0.13636979460716248, "learning_rate": 3.519035809814911e-05, "loss": 0.0206, "step": 98550 }, { "epoch": 1.9695455817113623, "grad_norm": 0.1191336065530777, "learning_rate": 3.518048794213251e-05, "loss": 0.0287, "step": 98560 }, { "epoch": 1.969745413852364, "grad_norm": 0.10220282524824142, "learning_rate": 3.517061841917825e-05, "loss": 0.0203, "step": 98570 }, { "epoch": 1.9699452459933657, "grad_norm": 0.1372644454240799, "learning_rate": 3.51607495297079e-05, "loss": 0.0241, "step": 98580 }, { "epoch": 1.970145078134367, "grad_norm": 0.13020119071006775, "learning_rate": 3.51508812741431e-05, "loss": 0.0277, "step": 98590 }, { "epoch": 1.9703449102753687, "grad_norm": 0.10537269711494446, "learning_rate": 3.5141013652905366e-05, "loss": 0.0226, "step": 98600 }, { "epoch": 1.9705447424163702, "grad_norm": 0.12651340663433075, "learning_rate": 3.5131146666416235e-05, "loss": 0.0261, "step": 98610 }, { "epoch": 1.9707445745573717, "grad_norm": 0.14072665572166443, "learning_rate": 3.5121280315097196e-05, "loss": 0.0331, "step": 98620 }, { "epoch": 1.9709444066983735, "grad_norm": 0.14436331391334534, "learning_rate": 3.511141459936972e-05, "loss": 0.0298, "step": 98630 }, { "epoch": 1.971144238839375, "grad_norm": 0.17087611556053162, "learning_rate": 3.510154951965527e-05, "loss": 0.0278, "step": 98640 }, { "epoch": 1.9713440709803765, "grad_norm": 0.2577835023403168, "learning_rate": 3.509168507637526e-05, "loss": 0.0304, "step": 98650 }, { "epoch": 1.9715439031213782, "grad_norm": 0.12072092294692993, "learning_rate": 3.5081821269951085e-05, "loss": 0.0205, "step": 98660 }, { "epoch": 1.9717437352623794, "grad_norm": 0.16349251568317413, "learning_rate": 3.507195810080407e-05, "loss": 0.0205, "step": 98670 }, { "epoch": 1.9719435674033812, "grad_norm": 0.1624760925769806, "learning_rate": 3.5062095569355595e-05, "loss": 0.0245, "step": 98680 }, { "epoch": 1.9721433995443827, "grad_norm": 0.09486126154661179, "learning_rate": 3.505223367602697e-05, "loss": 0.0328, "step": 98690 }, { "epoch": 1.9723432316853842, "grad_norm": 0.17646606266498566, "learning_rate": 3.504237242123945e-05, "loss": 0.0257, "step": 98700 }, { "epoch": 1.972543063826386, "grad_norm": 0.1231248751282692, "learning_rate": 3.503251180541431e-05, "loss": 0.0187, "step": 98710 }, { "epoch": 1.9727428959673874, "grad_norm": 0.1400601863861084, "learning_rate": 3.502265182897275e-05, "loss": 0.0226, "step": 98720 }, { "epoch": 1.9729427281083889, "grad_norm": 0.13432490825653076, "learning_rate": 3.5012792492336e-05, "loss": 0.0243, "step": 98730 }, { "epoch": 1.9731425602493906, "grad_norm": 0.1121673509478569, "learning_rate": 3.5002933795925236e-05, "loss": 0.032, "step": 98740 }, { "epoch": 1.973342392390392, "grad_norm": 0.1059279665350914, "learning_rate": 3.4993075740161574e-05, "loss": 0.0288, "step": 98750 }, { "epoch": 1.9735422245313936, "grad_norm": 0.11271683126688004, "learning_rate": 3.4983218325466145e-05, "loss": 0.0295, "step": 98760 }, { "epoch": 1.9737420566723953, "grad_norm": 0.10892771184444427, "learning_rate": 3.497336155226006e-05, "loss": 0.0207, "step": 98770 }, { "epoch": 1.9739418888133966, "grad_norm": 0.09786354005336761, "learning_rate": 3.496350542096435e-05, "loss": 0.0326, "step": 98780 }, { "epoch": 1.9741417209543983, "grad_norm": 0.13973736763000488, "learning_rate": 3.495364993200007e-05, "loss": 0.0264, "step": 98790 }, { "epoch": 1.9743415530953998, "grad_norm": 0.14100170135498047, "learning_rate": 3.494379508578823e-05, "loss": 0.0276, "step": 98800 }, { "epoch": 1.9745413852364013, "grad_norm": 0.13519971072673798, "learning_rate": 3.493394088274979e-05, "loss": 0.0265, "step": 98810 }, { "epoch": 1.974741217377403, "grad_norm": 0.10766036063432693, "learning_rate": 3.492408732330573e-05, "loss": 0.0351, "step": 98820 }, { "epoch": 1.9749410495184045, "grad_norm": 0.10029729455709457, "learning_rate": 3.4914234407876964e-05, "loss": 0.0238, "step": 98830 }, { "epoch": 1.975140881659406, "grad_norm": 0.14143186807632446, "learning_rate": 3.4904382136884395e-05, "loss": 0.0323, "step": 98840 }, { "epoch": 1.9753407138004078, "grad_norm": 0.1487167328596115, "learning_rate": 3.489453051074887e-05, "loss": 0.0275, "step": 98850 }, { "epoch": 1.9755405459414093, "grad_norm": 0.14635388553142548, "learning_rate": 3.488467952989126e-05, "loss": 0.0247, "step": 98860 }, { "epoch": 1.9757403780824108, "grad_norm": 0.12305012345314026, "learning_rate": 3.4874829194732386e-05, "loss": 0.0236, "step": 98870 }, { "epoch": 1.9759402102234125, "grad_norm": 0.11616234481334686, "learning_rate": 3.4864979505693017e-05, "loss": 0.0185, "step": 98880 }, { "epoch": 1.9761400423644138, "grad_norm": 0.13436459004878998, "learning_rate": 3.485513046319392e-05, "loss": 0.0206, "step": 98890 }, { "epoch": 1.9763398745054155, "grad_norm": 0.09982124716043472, "learning_rate": 3.4845282067655815e-05, "loss": 0.0231, "step": 98900 }, { "epoch": 1.976539706646417, "grad_norm": 0.1494988203048706, "learning_rate": 3.483543431949943e-05, "loss": 0.0298, "step": 98910 }, { "epoch": 1.9767395387874185, "grad_norm": 0.11312445998191833, "learning_rate": 3.482558721914544e-05, "loss": 0.027, "step": 98920 }, { "epoch": 1.9769393709284202, "grad_norm": 0.12199082970619202, "learning_rate": 3.481574076701448e-05, "loss": 0.0257, "step": 98930 }, { "epoch": 1.9771392030694217, "grad_norm": 0.11754739284515381, "learning_rate": 3.4805894963527166e-05, "loss": 0.0294, "step": 98940 }, { "epoch": 1.9773390352104232, "grad_norm": 0.12306072562932968, "learning_rate": 3.479604980910412e-05, "loss": 0.0327, "step": 98950 }, { "epoch": 1.977538867351425, "grad_norm": 0.1182904988527298, "learning_rate": 3.4786205304165895e-05, "loss": 0.0221, "step": 98960 }, { "epoch": 1.9777386994924262, "grad_norm": 0.10745014250278473, "learning_rate": 3.4776361449133035e-05, "loss": 0.0298, "step": 98970 }, { "epoch": 1.977938531633428, "grad_norm": 0.11970683187246323, "learning_rate": 3.476651824442606e-05, "loss": 0.0242, "step": 98980 }, { "epoch": 1.9781383637744296, "grad_norm": 0.08846494555473328, "learning_rate": 3.475667569046541e-05, "loss": 0.0272, "step": 98990 }, { "epoch": 1.978338195915431, "grad_norm": 0.13429801166057587, "learning_rate": 3.4746833787671585e-05, "loss": 0.0278, "step": 99000 }, { "epoch": 1.9785380280564326, "grad_norm": 0.1387045681476593, "learning_rate": 3.4736992536465005e-05, "loss": 0.0225, "step": 99010 }, { "epoch": 1.9787378601974341, "grad_norm": 0.14047807455062866, "learning_rate": 3.4727151937266054e-05, "loss": 0.0221, "step": 99020 }, { "epoch": 1.9789376923384356, "grad_norm": 0.1104697659611702, "learning_rate": 3.471731199049512e-05, "loss": 0.0298, "step": 99030 }, { "epoch": 1.9791375244794374, "grad_norm": 0.0903259664773941, "learning_rate": 3.470747269657255e-05, "loss": 0.0239, "step": 99040 }, { "epoch": 1.9793373566204389, "grad_norm": 0.1474352478981018, "learning_rate": 3.4697634055918646e-05, "loss": 0.0294, "step": 99050 }, { "epoch": 1.9795371887614404, "grad_norm": 0.11366408318281174, "learning_rate": 3.468779606895371e-05, "loss": 0.0279, "step": 99060 }, { "epoch": 1.979737020902442, "grad_norm": 0.10790611803531647, "learning_rate": 3.467795873609799e-05, "loss": 0.0294, "step": 99070 }, { "epoch": 1.9799368530434434, "grad_norm": 0.12654593586921692, "learning_rate": 3.466812205777172e-05, "loss": 0.0312, "step": 99080 }, { "epoch": 1.980136685184445, "grad_norm": 0.1302708089351654, "learning_rate": 3.465828603439513e-05, "loss": 0.0256, "step": 99090 }, { "epoch": 1.9803365173254466, "grad_norm": 0.12154140323400497, "learning_rate": 3.4648450666388354e-05, "loss": 0.0258, "step": 99100 }, { "epoch": 1.980536349466448, "grad_norm": 0.16706395149230957, "learning_rate": 3.463861595417158e-05, "loss": 0.033, "step": 99110 }, { "epoch": 1.9807361816074498, "grad_norm": 0.09337348490953445, "learning_rate": 3.4628781898164895e-05, "loss": 0.0203, "step": 99120 }, { "epoch": 1.9809360137484513, "grad_norm": 0.1119234710931778, "learning_rate": 3.461894849878843e-05, "loss": 0.0307, "step": 99130 }, { "epoch": 1.9811358458894528, "grad_norm": 0.12172941863536835, "learning_rate": 3.460911575646222e-05, "loss": 0.0222, "step": 99140 }, { "epoch": 1.9813356780304545, "grad_norm": 0.13596197962760925, "learning_rate": 3.459928367160631e-05, "loss": 0.0266, "step": 99150 }, { "epoch": 1.981535510171456, "grad_norm": 0.1513785421848297, "learning_rate": 3.458945224464073e-05, "loss": 0.0308, "step": 99160 }, { "epoch": 1.9817353423124575, "grad_norm": 0.14467011392116547, "learning_rate": 3.45796214759854e-05, "loss": 0.0304, "step": 99170 }, { "epoch": 1.9819351744534592, "grad_norm": 0.14516609907150269, "learning_rate": 3.4569791366060336e-05, "loss": 0.0294, "step": 99180 }, { "epoch": 1.9821350065944605, "grad_norm": 0.17015734314918518, "learning_rate": 3.455996191528545e-05, "loss": 0.0279, "step": 99190 }, { "epoch": 1.9823348387354622, "grad_norm": 0.1239011138677597, "learning_rate": 3.45501331240806e-05, "loss": 0.0294, "step": 99200 }, { "epoch": 1.9825346708764637, "grad_norm": 0.10606134682893753, "learning_rate": 3.4540304992865684e-05, "loss": 0.021, "step": 99210 }, { "epoch": 1.9827345030174652, "grad_norm": 0.1001628041267395, "learning_rate": 3.453047752206056e-05, "loss": 0.0292, "step": 99220 }, { "epoch": 1.982934335158467, "grad_norm": 0.1295737326145172, "learning_rate": 3.4520650712085e-05, "loss": 0.028, "step": 99230 }, { "epoch": 1.9831341672994685, "grad_norm": 0.1237354651093483, "learning_rate": 3.451082456335882e-05, "loss": 0.033, "step": 99240 }, { "epoch": 1.98333399944047, "grad_norm": 0.16456186771392822, "learning_rate": 3.450099907630175e-05, "loss": 0.0352, "step": 99250 }, { "epoch": 1.9835338315814717, "grad_norm": 0.1103677973151207, "learning_rate": 3.4491174251333515e-05, "loss": 0.0225, "step": 99260 }, { "epoch": 1.9837336637224732, "grad_norm": 0.11739824712276459, "learning_rate": 3.448135008887384e-05, "loss": 0.0272, "step": 99270 }, { "epoch": 1.9839334958634747, "grad_norm": 0.10747833549976349, "learning_rate": 3.4471526589342385e-05, "loss": 0.0307, "step": 99280 }, { "epoch": 1.9841333280044764, "grad_norm": 0.09583640098571777, "learning_rate": 3.446170375315879e-05, "loss": 0.0288, "step": 99290 }, { "epoch": 1.9843331601454777, "grad_norm": 0.1065579429268837, "learning_rate": 3.445188158074263e-05, "loss": 0.0285, "step": 99300 }, { "epoch": 1.9845329922864794, "grad_norm": 0.11056586354970932, "learning_rate": 3.444206007251356e-05, "loss": 0.0296, "step": 99310 }, { "epoch": 1.984732824427481, "grad_norm": 0.1143023818731308, "learning_rate": 3.44322392288911e-05, "loss": 0.0264, "step": 99320 }, { "epoch": 1.9849326565684824, "grad_norm": 0.10419359803199768, "learning_rate": 3.4422419050294787e-05, "loss": 0.0242, "step": 99330 }, { "epoch": 1.985132488709484, "grad_norm": 0.16053876280784607, "learning_rate": 3.4412599537144094e-05, "loss": 0.0255, "step": 99340 }, { "epoch": 1.9853323208504856, "grad_norm": 0.1050700694322586, "learning_rate": 3.440278068985852e-05, "loss": 0.0285, "step": 99350 }, { "epoch": 1.985532152991487, "grad_norm": 0.12919652462005615, "learning_rate": 3.439296250885752e-05, "loss": 0.0277, "step": 99360 }, { "epoch": 1.9857319851324888, "grad_norm": 0.13060171902179718, "learning_rate": 3.4383144994560465e-05, "loss": 0.0337, "step": 99370 }, { "epoch": 1.98593181727349, "grad_norm": 0.09265793859958649, "learning_rate": 3.43733281473868e-05, "loss": 0.0214, "step": 99380 }, { "epoch": 1.9861316494144918, "grad_norm": 0.12237393856048584, "learning_rate": 3.436351196775581e-05, "loss": 0.0221, "step": 99390 }, { "epoch": 1.9863314815554935, "grad_norm": 0.12807106971740723, "learning_rate": 3.435369645608688e-05, "loss": 0.0295, "step": 99400 }, { "epoch": 1.9865313136964948, "grad_norm": 0.11404718458652496, "learning_rate": 3.4343881612799304e-05, "loss": 0.034, "step": 99410 }, { "epoch": 1.9867311458374965, "grad_norm": 0.09273435920476913, "learning_rate": 3.4334067438312334e-05, "loss": 0.0451, "step": 99420 }, { "epoch": 1.986930977978498, "grad_norm": 0.09893213957548141, "learning_rate": 3.432425393304522e-05, "loss": 0.0272, "step": 99430 }, { "epoch": 1.9871308101194995, "grad_norm": 0.12496412545442581, "learning_rate": 3.4314441097417195e-05, "loss": 0.0374, "step": 99440 }, { "epoch": 1.9873306422605013, "grad_norm": 0.12569326162338257, "learning_rate": 3.430462893184742e-05, "loss": 0.0255, "step": 99450 }, { "epoch": 1.9875304744015028, "grad_norm": 0.11080628633499146, "learning_rate": 3.429481743675508e-05, "loss": 0.0198, "step": 99460 }, { "epoch": 1.9877303065425043, "grad_norm": 0.11307752877473831, "learning_rate": 3.428500661255927e-05, "loss": 0.0278, "step": 99470 }, { "epoch": 1.987930138683506, "grad_norm": 0.11673557013273239, "learning_rate": 3.4275196459679104e-05, "loss": 0.0306, "step": 99480 }, { "epoch": 1.9881299708245073, "grad_norm": 0.10794967412948608, "learning_rate": 3.4265386978533675e-05, "loss": 0.026, "step": 99490 }, { "epoch": 1.988329802965509, "grad_norm": 0.11103113740682602, "learning_rate": 3.4255578169542e-05, "loss": 0.02, "step": 99500 }, { "epoch": 1.9885296351065105, "grad_norm": 0.12711021304130554, "learning_rate": 3.424577003312311e-05, "loss": 0.0341, "step": 99510 }, { "epoch": 1.988729467247512, "grad_norm": 0.13610537350177765, "learning_rate": 3.423596256969596e-05, "loss": 0.0276, "step": 99520 }, { "epoch": 1.9889292993885137, "grad_norm": 0.20606866478919983, "learning_rate": 3.422615577967955e-05, "loss": 0.0347, "step": 99530 }, { "epoch": 1.9891291315295152, "grad_norm": 0.12105660140514374, "learning_rate": 3.421634966349279e-05, "loss": 0.0262, "step": 99540 }, { "epoch": 1.9893289636705167, "grad_norm": 0.11169693619012833, "learning_rate": 3.4206544221554565e-05, "loss": 0.0272, "step": 99550 }, { "epoch": 1.9895287958115184, "grad_norm": 0.1284799724817276, "learning_rate": 3.419673945428377e-05, "loss": 0.0223, "step": 99560 }, { "epoch": 1.98972862795252, "grad_norm": 0.11570589989423752, "learning_rate": 3.418693536209922e-05, "loss": 0.0304, "step": 99570 }, { "epoch": 1.9899284600935214, "grad_norm": 0.11538831144571304, "learning_rate": 3.417713194541974e-05, "loss": 0.0239, "step": 99580 }, { "epoch": 1.9901282922345231, "grad_norm": 0.13187937438488007, "learning_rate": 3.416732920466413e-05, "loss": 0.0275, "step": 99590 }, { "epoch": 1.9903281243755244, "grad_norm": 0.10876524448394775, "learning_rate": 3.415752714025112e-05, "loss": 0.0248, "step": 99600 }, { "epoch": 1.9905279565165261, "grad_norm": 0.11723756045103073, "learning_rate": 3.4147725752599446e-05, "loss": 0.0321, "step": 99610 }, { "epoch": 1.9907277886575276, "grad_norm": 0.0982871726155281, "learning_rate": 3.413792504212782e-05, "loss": 0.0386, "step": 99620 }, { "epoch": 1.9909276207985291, "grad_norm": 0.09324069321155548, "learning_rate": 3.412812500925488e-05, "loss": 0.029, "step": 99630 }, { "epoch": 1.9911274529395309, "grad_norm": 0.12885300815105438, "learning_rate": 3.411832565439931e-05, "loss": 0.0265, "step": 99640 }, { "epoch": 1.9913272850805324, "grad_norm": 0.10867565125226974, "learning_rate": 3.4108526977979665e-05, "loss": 0.0322, "step": 99650 }, { "epoch": 1.9915271172215339, "grad_norm": 0.10733629763126373, "learning_rate": 3.409872898041455e-05, "loss": 0.0267, "step": 99660 }, { "epoch": 1.9917269493625356, "grad_norm": 0.15436942875385284, "learning_rate": 3.408893166212254e-05, "loss": 0.0272, "step": 99670 }, { "epoch": 1.991926781503537, "grad_norm": 0.15885715186595917, "learning_rate": 3.407913502352214e-05, "loss": 0.0314, "step": 99680 }, { "epoch": 1.9921266136445386, "grad_norm": 0.14334434270858765, "learning_rate": 3.406933906503184e-05, "loss": 0.0294, "step": 99690 }, { "epoch": 1.9923264457855403, "grad_norm": 0.14158695936203003, "learning_rate": 3.405954378707008e-05, "loss": 0.0298, "step": 99700 }, { "epoch": 1.9925262779265416, "grad_norm": 0.10894272476434708, "learning_rate": 3.404974919005537e-05, "loss": 0.0283, "step": 99710 }, { "epoch": 1.9927261100675433, "grad_norm": 0.11792664229869843, "learning_rate": 3.403995527440605e-05, "loss": 0.0245, "step": 99720 }, { "epoch": 1.9929259422085448, "grad_norm": 0.11880970746278763, "learning_rate": 3.4030162040540536e-05, "loss": 0.0251, "step": 99730 }, { "epoch": 1.9931257743495463, "grad_norm": 0.14438538253307343, "learning_rate": 3.402036948887715e-05, "loss": 0.0327, "step": 99740 }, { "epoch": 1.993325606490548, "grad_norm": 0.1271386295557022, "learning_rate": 3.4010577619834214e-05, "loss": 0.0205, "step": 99750 }, { "epoch": 1.9935254386315495, "grad_norm": 0.1342691332101822, "learning_rate": 3.400078643383005e-05, "loss": 0.0381, "step": 99760 }, { "epoch": 1.993725270772551, "grad_norm": 0.12109018117189407, "learning_rate": 3.399099593128289e-05, "loss": 0.0233, "step": 99770 }, { "epoch": 1.9939251029135527, "grad_norm": 0.11692158877849579, "learning_rate": 3.3981206112610986e-05, "loss": 0.0266, "step": 99780 }, { "epoch": 1.994124935054554, "grad_norm": 0.12525279819965363, "learning_rate": 3.39714169782325e-05, "loss": 0.0198, "step": 99790 }, { "epoch": 1.9943247671955557, "grad_norm": 0.13753578066825867, "learning_rate": 3.3961628528565656e-05, "loss": 0.0234, "step": 99800 }, { "epoch": 1.9945245993365575, "grad_norm": 0.1266522854566574, "learning_rate": 3.395184076402858e-05, "loss": 0.0363, "step": 99810 }, { "epoch": 1.9947244314775587, "grad_norm": 0.11090859770774841, "learning_rate": 3.394205368503938e-05, "loss": 0.0273, "step": 99820 }, { "epoch": 1.9949242636185605, "grad_norm": 0.10900397598743439, "learning_rate": 3.393226729201615e-05, "loss": 0.036, "step": 99830 }, { "epoch": 1.995124095759562, "grad_norm": 0.09110521525144577, "learning_rate": 3.392248158537693e-05, "loss": 0.0246, "step": 99840 }, { "epoch": 1.9953239279005635, "grad_norm": 0.16410940885543823, "learning_rate": 3.391269656553977e-05, "loss": 0.0349, "step": 99850 }, { "epoch": 1.9955237600415652, "grad_norm": 0.13634410500526428, "learning_rate": 3.3902912232922666e-05, "loss": 0.0226, "step": 99860 }, { "epoch": 1.9957235921825667, "grad_norm": 0.1426108479499817, "learning_rate": 3.389312858794358e-05, "loss": 0.0296, "step": 99870 }, { "epoch": 1.9959234243235682, "grad_norm": 0.13071513175964355, "learning_rate": 3.3883345631020434e-05, "loss": 0.0266, "step": 99880 }, { "epoch": 1.99612325646457, "grad_norm": 0.1424422413110733, "learning_rate": 3.387356336257118e-05, "loss": 0.0212, "step": 99890 }, { "epoch": 1.9963230886055712, "grad_norm": 0.09606808423995972, "learning_rate": 3.386378178301366e-05, "loss": 0.0306, "step": 99900 }, { "epoch": 1.996522920746573, "grad_norm": 0.123298279941082, "learning_rate": 3.3854000892765745e-05, "loss": 0.0362, "step": 99910 }, { "epoch": 1.9967227528875744, "grad_norm": 0.13387709856033325, "learning_rate": 3.384422069224525e-05, "loss": 0.0219, "step": 99920 }, { "epoch": 1.996922585028576, "grad_norm": 0.11140964925289154, "learning_rate": 3.383444118186995e-05, "loss": 0.0288, "step": 99930 }, { "epoch": 1.9971224171695776, "grad_norm": 0.13305708765983582, "learning_rate": 3.3824662362057644e-05, "loss": 0.0279, "step": 99940 }, { "epoch": 1.9973222493105791, "grad_norm": 0.12451651692390442, "learning_rate": 3.3814884233226046e-05, "loss": 0.0243, "step": 99950 }, { "epoch": 1.9975220814515806, "grad_norm": 0.14457160234451294, "learning_rate": 3.380510679579287e-05, "loss": 0.0223, "step": 99960 }, { "epoch": 1.9977219135925823, "grad_norm": 0.11855961382389069, "learning_rate": 3.379533005017575e-05, "loss": 0.0289, "step": 99970 }, { "epoch": 1.9979217457335838, "grad_norm": 0.1839025616645813, "learning_rate": 3.3785553996792386e-05, "loss": 0.032, "step": 99980 }, { "epoch": 1.9981215778745853, "grad_norm": 0.1436910778284073, "learning_rate": 3.3775778636060367e-05, "loss": 0.0377, "step": 99990 }, { "epoch": 1.998321410015587, "grad_norm": 0.13707335293293, "learning_rate": 3.3766003968397274e-05, "loss": 0.0275, "step": 100000 }, { "epoch": 1.9985212421565883, "grad_norm": 0.1323830783367157, "learning_rate": 3.3756229994220676e-05, "loss": 0.0233, "step": 100010 }, { "epoch": 1.99872107429759, "grad_norm": 0.16079291701316833, "learning_rate": 3.374645671394808e-05, "loss": 0.0304, "step": 100020 }, { "epoch": 1.9989209064385915, "grad_norm": 0.09000439196825027, "learning_rate": 3.3736684127997e-05, "loss": 0.0244, "step": 100030 }, { "epoch": 1.999120738579593, "grad_norm": 0.11951445043087006, "learning_rate": 3.37269122367849e-05, "loss": 0.0236, "step": 100040 }, { "epoch": 1.9993205707205948, "grad_norm": 0.13039886951446533, "learning_rate": 3.371714104072921e-05, "loss": 0.0285, "step": 100050 }, { "epoch": 1.9995204028615963, "grad_norm": 0.12472414970397949, "learning_rate": 3.370737054024734e-05, "loss": 0.0285, "step": 100060 }, { "epoch": 1.9997202350025978, "grad_norm": 0.11052314192056656, "learning_rate": 3.369760073575667e-05, "loss": 0.0322, "step": 100070 }, { "epoch": 1.9999200671435995, "grad_norm": 0.08316770941019058, "learning_rate": 3.368783162767456e-05, "loss": 0.0207, "step": 100080 }, { "epoch": 2.0001198992846008, "grad_norm": 0.11013966053724289, "learning_rate": 3.367806321641832e-05, "loss": 0.0325, "step": 100090 }, { "epoch": 2.0003197314256025, "grad_norm": 0.13552136719226837, "learning_rate": 3.3668295502405216e-05, "loss": 0.025, "step": 100100 }, { "epoch": 2.000519563566604, "grad_norm": 0.14515748620033264, "learning_rate": 3.365852848605252e-05, "loss": 0.0289, "step": 100110 }, { "epoch": 2.0007193957076055, "grad_norm": 0.10241696238517761, "learning_rate": 3.364876216777748e-05, "loss": 0.0266, "step": 100120 }, { "epoch": 2.000919227848607, "grad_norm": 0.15911266207695007, "learning_rate": 3.363899654799729e-05, "loss": 0.0243, "step": 100130 }, { "epoch": 2.001119059989609, "grad_norm": 0.16600753366947174, "learning_rate": 3.362923162712911e-05, "loss": 0.0323, "step": 100140 }, { "epoch": 2.00131889213061, "grad_norm": 0.15864509344100952, "learning_rate": 3.361946740559006e-05, "loss": 0.0366, "step": 100150 }, { "epoch": 2.001518724271612, "grad_norm": 0.13112632930278778, "learning_rate": 3.3609703883797294e-05, "loss": 0.027, "step": 100160 }, { "epoch": 2.001718556412613, "grad_norm": 0.15140028297901154, "learning_rate": 3.3599941062167857e-05, "loss": 0.0216, "step": 100170 }, { "epoch": 2.001918388553615, "grad_norm": 0.13621897995471954, "learning_rate": 3.359017894111882e-05, "loss": 0.0249, "step": 100180 }, { "epoch": 2.0021182206946166, "grad_norm": 0.10237428545951843, "learning_rate": 3.3580417521067186e-05, "loss": 0.0194, "step": 100190 }, { "epoch": 2.002318052835618, "grad_norm": 0.10809320211410522, "learning_rate": 3.3570656802429934e-05, "loss": 0.0292, "step": 100200 }, { "epoch": 2.0025178849766196, "grad_norm": 0.11817361414432526, "learning_rate": 3.3560896785624077e-05, "loss": 0.0361, "step": 100210 }, { "epoch": 2.0027177171176214, "grad_norm": 0.1601356863975525, "learning_rate": 3.355113747106648e-05, "loss": 0.0301, "step": 100220 }, { "epoch": 2.0029175492586226, "grad_norm": 0.12272724509239197, "learning_rate": 3.35413788591741e-05, "loss": 0.0261, "step": 100230 }, { "epoch": 2.0031173813996244, "grad_norm": 0.12129464745521545, "learning_rate": 3.3531620950363754e-05, "loss": 0.0318, "step": 100240 }, { "epoch": 2.003317213540626, "grad_norm": 0.1261897087097168, "learning_rate": 3.352186374505231e-05, "loss": 0.0204, "step": 100250 }, { "epoch": 2.0035170456816274, "grad_norm": 0.08553609997034073, "learning_rate": 3.351210724365659e-05, "loss": 0.0244, "step": 100260 }, { "epoch": 2.003716877822629, "grad_norm": 0.10372645407915115, "learning_rate": 3.350235144659335e-05, "loss": 0.0256, "step": 100270 }, { "epoch": 2.0039167099636304, "grad_norm": 0.0832701027393341, "learning_rate": 3.349259635427936e-05, "loss": 0.0239, "step": 100280 }, { "epoch": 2.004116542104632, "grad_norm": 0.14026425778865814, "learning_rate": 3.3482841967131305e-05, "loss": 0.0285, "step": 100290 }, { "epoch": 2.004316374245634, "grad_norm": 0.13472694158554077, "learning_rate": 3.347308828556591e-05, "loss": 0.0322, "step": 100300 }, { "epoch": 2.004516206386635, "grad_norm": 0.13615386188030243, "learning_rate": 3.3463335309999847e-05, "loss": 0.0275, "step": 100310 }, { "epoch": 2.004716038527637, "grad_norm": 0.10649415850639343, "learning_rate": 3.345358304084969e-05, "loss": 0.0292, "step": 100320 }, { "epoch": 2.0049158706686385, "grad_norm": 0.12863916158676147, "learning_rate": 3.344383147853207e-05, "loss": 0.0285, "step": 100330 }, { "epoch": 2.00511570280964, "grad_norm": 0.15200959146022797, "learning_rate": 3.343408062346357e-05, "loss": 0.0231, "step": 100340 }, { "epoch": 2.0053155349506415, "grad_norm": 0.17304490506649017, "learning_rate": 3.3424330476060705e-05, "loss": 0.0287, "step": 100350 }, { "epoch": 2.0055153670916432, "grad_norm": 0.11668583750724792, "learning_rate": 3.341458103674e-05, "loss": 0.0249, "step": 100360 }, { "epoch": 2.0057151992326445, "grad_norm": 0.13024041056632996, "learning_rate": 3.340483230591792e-05, "loss": 0.022, "step": 100370 }, { "epoch": 2.0059150313736462, "grad_norm": 0.12324460595846176, "learning_rate": 3.33950842840109e-05, "loss": 0.0239, "step": 100380 }, { "epoch": 2.0061148635146475, "grad_norm": 0.10706109553575516, "learning_rate": 3.33853369714354e-05, "loss": 0.0299, "step": 100390 }, { "epoch": 2.0063146956556492, "grad_norm": 0.13396114110946655, "learning_rate": 3.3375590368607765e-05, "loss": 0.0302, "step": 100400 }, { "epoch": 2.006514527796651, "grad_norm": 0.11393119394779205, "learning_rate": 3.336584447594438e-05, "loss": 0.0304, "step": 100410 }, { "epoch": 2.0067143599376522, "grad_norm": 0.1411871761083603, "learning_rate": 3.3356099293861555e-05, "loss": 0.0343, "step": 100420 }, { "epoch": 2.006914192078654, "grad_norm": 0.1331975907087326, "learning_rate": 3.3346354822775595e-05, "loss": 0.0267, "step": 100430 }, { "epoch": 2.0071140242196557, "grad_norm": 0.13005821406841278, "learning_rate": 3.3336611063102774e-05, "loss": 0.0276, "step": 100440 }, { "epoch": 2.007313856360657, "grad_norm": 0.10383490473031998, "learning_rate": 3.33268680152593e-05, "loss": 0.0283, "step": 100450 }, { "epoch": 2.0075136885016587, "grad_norm": 0.08141400665044785, "learning_rate": 3.331712567966141e-05, "loss": 0.0222, "step": 100460 }, { "epoch": 2.00771352064266, "grad_norm": 0.1131652221083641, "learning_rate": 3.330738405672524e-05, "loss": 0.023, "step": 100470 }, { "epoch": 2.0079133527836617, "grad_norm": 0.1394975781440735, "learning_rate": 3.329764314686698e-05, "loss": 0.0247, "step": 100480 }, { "epoch": 2.0081131849246634, "grad_norm": 0.1286044716835022, "learning_rate": 3.328790295050271e-05, "loss": 0.0283, "step": 100490 }, { "epoch": 2.0083130170656647, "grad_norm": 0.15186473727226257, "learning_rate": 3.3278163468048534e-05, "loss": 0.0257, "step": 100500 }, { "epoch": 2.0085128492066664, "grad_norm": 0.0975530818104744, "learning_rate": 3.3268424699920476e-05, "loss": 0.0346, "step": 100510 }, { "epoch": 2.008712681347668, "grad_norm": 0.14679992198944092, "learning_rate": 3.325868664653459e-05, "loss": 0.0215, "step": 100520 }, { "epoch": 2.0089125134886694, "grad_norm": 0.12269365787506104, "learning_rate": 3.324894930830687e-05, "loss": 0.038, "step": 100530 }, { "epoch": 2.009112345629671, "grad_norm": 0.1402457058429718, "learning_rate": 3.323921268565325e-05, "loss": 0.0247, "step": 100540 }, { "epoch": 2.009312177770673, "grad_norm": 0.12807098031044006, "learning_rate": 3.3229476778989664e-05, "loss": 0.0223, "step": 100550 }, { "epoch": 2.009512009911674, "grad_norm": 0.10481058061122894, "learning_rate": 3.321974158873204e-05, "loss": 0.0266, "step": 100560 }, { "epoch": 2.009711842052676, "grad_norm": 0.16357280313968658, "learning_rate": 3.3210007115296215e-05, "loss": 0.0232, "step": 100570 }, { "epoch": 2.009911674193677, "grad_norm": 0.13124123215675354, "learning_rate": 3.320027335909806e-05, "loss": 0.037, "step": 100580 }, { "epoch": 2.010111506334679, "grad_norm": 0.1587737798690796, "learning_rate": 3.3190540320553354e-05, "loss": 0.0226, "step": 100590 }, { "epoch": 2.0103113384756806, "grad_norm": 0.10466986894607544, "learning_rate": 3.318080800007787e-05, "loss": 0.018, "step": 100600 }, { "epoch": 2.010511170616682, "grad_norm": 0.14762955904006958, "learning_rate": 3.31710763980874e-05, "loss": 0.0289, "step": 100610 }, { "epoch": 2.0107110027576836, "grad_norm": 0.11718881875276566, "learning_rate": 3.316134551499761e-05, "loss": 0.0265, "step": 100620 }, { "epoch": 2.0109108348986853, "grad_norm": 0.12386693805456161, "learning_rate": 3.315161535122422e-05, "loss": 0.0291, "step": 100630 }, { "epoch": 2.0111106670396866, "grad_norm": 0.11207111179828644, "learning_rate": 3.3141885907182845e-05, "loss": 0.0255, "step": 100640 }, { "epoch": 2.0113104991806883, "grad_norm": 0.13158684968948364, "learning_rate": 3.313215718328915e-05, "loss": 0.0277, "step": 100650 }, { "epoch": 2.01151033132169, "grad_norm": 0.10498719662427902, "learning_rate": 3.312242917995872e-05, "loss": 0.0255, "step": 100660 }, { "epoch": 2.0117101634626913, "grad_norm": 0.141363725066185, "learning_rate": 3.311270189760711e-05, "loss": 0.0274, "step": 100670 }, { "epoch": 2.011909995603693, "grad_norm": 0.15933112800121307, "learning_rate": 3.310297533664985e-05, "loss": 0.0254, "step": 100680 }, { "epoch": 2.0121098277446943, "grad_norm": 0.1336999088525772, "learning_rate": 3.3093249497502434e-05, "loss": 0.0224, "step": 100690 }, { "epoch": 2.012309659885696, "grad_norm": 0.12468994408845901, "learning_rate": 3.308352438058034e-05, "loss": 0.0238, "step": 100700 }, { "epoch": 2.0125094920266977, "grad_norm": 0.0983404666185379, "learning_rate": 3.307379998629903e-05, "loss": 0.0339, "step": 100710 }, { "epoch": 2.012709324167699, "grad_norm": 0.14415575563907623, "learning_rate": 3.306407631507387e-05, "loss": 0.0344, "step": 100720 }, { "epoch": 2.0129091563087007, "grad_norm": 0.13561676442623138, "learning_rate": 3.3054353367320254e-05, "loss": 0.0268, "step": 100730 }, { "epoch": 2.0131089884497024, "grad_norm": 0.1401376724243164, "learning_rate": 3.3044631143453544e-05, "loss": 0.0289, "step": 100740 }, { "epoch": 2.0133088205907037, "grad_norm": 0.13203009963035583, "learning_rate": 3.3034909643889045e-05, "loss": 0.0281, "step": 100750 }, { "epoch": 2.0135086527317054, "grad_norm": 0.11812196671962738, "learning_rate": 3.302518886904204e-05, "loss": 0.026, "step": 100760 }, { "epoch": 2.0137084848727067, "grad_norm": 0.1287158876657486, "learning_rate": 3.3015468819327775e-05, "loss": 0.0338, "step": 100770 }, { "epoch": 2.0139083170137084, "grad_norm": 0.11040204763412476, "learning_rate": 3.300574949516148e-05, "loss": 0.0272, "step": 100780 }, { "epoch": 2.01410814915471, "grad_norm": 0.1094779223203659, "learning_rate": 3.299603089695835e-05, "loss": 0.0227, "step": 100790 }, { "epoch": 2.0143079812957114, "grad_norm": 0.10299850255250931, "learning_rate": 3.298631302513354e-05, "loss": 0.0221, "step": 100800 }, { "epoch": 2.014507813436713, "grad_norm": 0.11335291713476181, "learning_rate": 3.29765958801022e-05, "loss": 0.0243, "step": 100810 }, { "epoch": 2.014707645577715, "grad_norm": 0.12590549886226654, "learning_rate": 3.2966879462279385e-05, "loss": 0.0286, "step": 100820 }, { "epoch": 2.014907477718716, "grad_norm": 0.13149145245552063, "learning_rate": 3.29571637720802e-05, "loss": 0.028, "step": 100830 }, { "epoch": 2.015107309859718, "grad_norm": 0.10414620488882065, "learning_rate": 3.294744880991968e-05, "loss": 0.0284, "step": 100840 }, { "epoch": 2.0153071420007196, "grad_norm": 0.12300282716751099, "learning_rate": 3.2937734576212805e-05, "loss": 0.03, "step": 100850 }, { "epoch": 2.015506974141721, "grad_norm": 0.11173484474420547, "learning_rate": 3.292802107137458e-05, "loss": 0.021, "step": 100860 }, { "epoch": 2.0157068062827226, "grad_norm": 0.12811820209026337, "learning_rate": 3.291830829581991e-05, "loss": 0.0315, "step": 100870 }, { "epoch": 2.015906638423724, "grad_norm": 0.12825459241867065, "learning_rate": 3.290859624996374e-05, "loss": 0.0318, "step": 100880 }, { "epoch": 2.0161064705647256, "grad_norm": 0.11201226711273193, "learning_rate": 3.289888493422095e-05, "loss": 0.0246, "step": 100890 }, { "epoch": 2.0163063027057273, "grad_norm": 0.13188257813453674, "learning_rate": 3.2889174349006366e-05, "loss": 0.0253, "step": 100900 }, { "epoch": 2.0165061348467286, "grad_norm": 0.15390340983867645, "learning_rate": 3.287946449473482e-05, "loss": 0.0304, "step": 100910 }, { "epoch": 2.0167059669877303, "grad_norm": 0.11978430300951004, "learning_rate": 3.2869755371821096e-05, "loss": 0.028, "step": 100920 }, { "epoch": 2.016905799128732, "grad_norm": 0.2045450210571289, "learning_rate": 3.286004698067998e-05, "loss": 0.0262, "step": 100930 }, { "epoch": 2.0171056312697333, "grad_norm": 0.14364008605480194, "learning_rate": 3.285033932172615e-05, "loss": 0.026, "step": 100940 }, { "epoch": 2.017305463410735, "grad_norm": 0.13706809282302856, "learning_rate": 3.2840632395374325e-05, "loss": 0.0329, "step": 100950 }, { "epoch": 2.0175052955517367, "grad_norm": 0.11814413964748383, "learning_rate": 3.283092620203915e-05, "loss": 0.0166, "step": 100960 }, { "epoch": 2.017705127692738, "grad_norm": 0.20214277505874634, "learning_rate": 3.2821220742135274e-05, "loss": 0.0278, "step": 100970 }, { "epoch": 2.0179049598337397, "grad_norm": 0.11903735995292664, "learning_rate": 3.28115160160773e-05, "loss": 0.0353, "step": 100980 }, { "epoch": 2.018104791974741, "grad_norm": 0.15664160251617432, "learning_rate": 3.2801812024279774e-05, "loss": 0.0239, "step": 100990 }, { "epoch": 2.0183046241157427, "grad_norm": 0.12171721458435059, "learning_rate": 3.2792108767157235e-05, "loss": 0.0299, "step": 101000 }, { "epoch": 2.0185044562567445, "grad_norm": 0.1502026915550232, "learning_rate": 3.278240624512422e-05, "loss": 0.0259, "step": 101010 }, { "epoch": 2.0187042883977457, "grad_norm": 0.1081671342253685, "learning_rate": 3.2772704458595174e-05, "loss": 0.0266, "step": 101020 }, { "epoch": 2.0189041205387475, "grad_norm": 0.13322483003139496, "learning_rate": 3.2763003407984545e-05, "loss": 0.0278, "step": 101030 }, { "epoch": 2.019103952679749, "grad_norm": 0.16236315667629242, "learning_rate": 3.275330309370674e-05, "loss": 0.0233, "step": 101040 }, { "epoch": 2.0193037848207505, "grad_norm": 0.12278610467910767, "learning_rate": 3.2743603516176136e-05, "loss": 0.0315, "step": 101050 }, { "epoch": 2.019503616961752, "grad_norm": 0.10255242139101028, "learning_rate": 3.2733904675807107e-05, "loss": 0.0278, "step": 101060 }, { "epoch": 2.019703449102754, "grad_norm": 0.19561231136322021, "learning_rate": 3.272420657301394e-05, "loss": 0.0226, "step": 101070 }, { "epoch": 2.019903281243755, "grad_norm": 0.12410695850849152, "learning_rate": 3.271450920821094e-05, "loss": 0.0308, "step": 101080 }, { "epoch": 2.020103113384757, "grad_norm": 0.13489842414855957, "learning_rate": 3.270481258181234e-05, "loss": 0.0217, "step": 101090 }, { "epoch": 2.020302945525758, "grad_norm": 0.1204376146197319, "learning_rate": 3.269511669423237e-05, "loss": 0.018, "step": 101100 }, { "epoch": 2.02050277766676, "grad_norm": 0.09937344491481781, "learning_rate": 3.2685421545885244e-05, "loss": 0.019, "step": 101110 }, { "epoch": 2.0207026098077616, "grad_norm": 0.1128377914428711, "learning_rate": 3.2675727137185094e-05, "loss": 0.037, "step": 101120 }, { "epoch": 2.020902441948763, "grad_norm": 0.12750503420829773, "learning_rate": 3.2666033468546065e-05, "loss": 0.0316, "step": 101130 }, { "epoch": 2.0211022740897646, "grad_norm": 0.09566736966371536, "learning_rate": 3.2656340540382216e-05, "loss": 0.0211, "step": 101140 }, { "epoch": 2.0213021062307663, "grad_norm": 0.16878367960453033, "learning_rate": 3.264664835310766e-05, "loss": 0.0264, "step": 101150 }, { "epoch": 2.0215019383717676, "grad_norm": 0.12215764075517654, "learning_rate": 3.2636956907136405e-05, "loss": 0.0283, "step": 101160 }, { "epoch": 2.0217017705127693, "grad_norm": 0.1063099205493927, "learning_rate": 3.262726620288244e-05, "loss": 0.0225, "step": 101170 }, { "epoch": 2.0219016026537706, "grad_norm": 0.10921914130449295, "learning_rate": 3.261757624075975e-05, "loss": 0.0277, "step": 101180 }, { "epoch": 2.0221014347947723, "grad_norm": 0.10382995754480362, "learning_rate": 3.2607887021182296e-05, "loss": 0.0241, "step": 101190 }, { "epoch": 2.022301266935774, "grad_norm": 0.0802059918642044, "learning_rate": 3.2598198544563936e-05, "loss": 0.0222, "step": 101200 }, { "epoch": 2.0225010990767753, "grad_norm": 0.17913952469825745, "learning_rate": 3.258851081131857e-05, "loss": 0.0237, "step": 101210 }, { "epoch": 2.022700931217777, "grad_norm": 0.11388864368200302, "learning_rate": 3.257882382186004e-05, "loss": 0.0274, "step": 101220 }, { "epoch": 2.0229007633587788, "grad_norm": 0.09976956993341446, "learning_rate": 3.2569137576602135e-05, "loss": 0.0269, "step": 101230 }, { "epoch": 2.02310059549978, "grad_norm": 0.12559117376804352, "learning_rate": 3.2559452075958677e-05, "loss": 0.0303, "step": 101240 }, { "epoch": 2.0233004276407818, "grad_norm": 0.1268850713968277, "learning_rate": 3.254976732034337e-05, "loss": 0.0243, "step": 101250 }, { "epoch": 2.0235002597817835, "grad_norm": 0.11310841888189316, "learning_rate": 3.2540083310169964e-05, "loss": 0.0253, "step": 101260 }, { "epoch": 2.0237000919227848, "grad_norm": 0.15184035897254944, "learning_rate": 3.25304000458521e-05, "loss": 0.0292, "step": 101270 }, { "epoch": 2.0238999240637865, "grad_norm": 0.11889611184597015, "learning_rate": 3.252071752780346e-05, "loss": 0.0227, "step": 101280 }, { "epoch": 2.0240997562047878, "grad_norm": 0.11273147910833359, "learning_rate": 3.2511035756437674e-05, "loss": 0.0281, "step": 101290 }, { "epoch": 2.0242995883457895, "grad_norm": 0.11358530074357986, "learning_rate": 3.25013547321683e-05, "loss": 0.029, "step": 101300 }, { "epoch": 2.024499420486791, "grad_norm": 0.1246873065829277, "learning_rate": 3.249167445540892e-05, "loss": 0.0284, "step": 101310 }, { "epoch": 2.0246992526277925, "grad_norm": 0.13474203646183014, "learning_rate": 3.2481994926573e-05, "loss": 0.0322, "step": 101320 }, { "epoch": 2.024899084768794, "grad_norm": 0.13796186447143555, "learning_rate": 3.247231614607411e-05, "loss": 0.0259, "step": 101330 }, { "epoch": 2.025098916909796, "grad_norm": 0.11010104417800903, "learning_rate": 3.246263811432567e-05, "loss": 0.03, "step": 101340 }, { "epoch": 2.025298749050797, "grad_norm": 0.1315513551235199, "learning_rate": 3.245296083174112e-05, "loss": 0.0258, "step": 101350 }, { "epoch": 2.025498581191799, "grad_norm": 0.13767704367637634, "learning_rate": 3.244328429873383e-05, "loss": 0.0237, "step": 101360 }, { "epoch": 2.0256984133328007, "grad_norm": 0.12156423926353455, "learning_rate": 3.243360851571719e-05, "loss": 0.0265, "step": 101370 }, { "epoch": 2.025898245473802, "grad_norm": 0.155417799949646, "learning_rate": 3.2423933483104527e-05, "loss": 0.0309, "step": 101380 }, { "epoch": 2.0260980776148036, "grad_norm": 0.18624907732009888, "learning_rate": 3.241425920130914e-05, "loss": 0.0363, "step": 101390 }, { "epoch": 2.026297909755805, "grad_norm": 0.12435514479875565, "learning_rate": 3.24045856707443e-05, "loss": 0.0216, "step": 101400 }, { "epoch": 2.0264977418968066, "grad_norm": 0.12107831239700317, "learning_rate": 3.239491289182321e-05, "loss": 0.0262, "step": 101410 }, { "epoch": 2.0266975740378084, "grad_norm": 0.09505094587802887, "learning_rate": 3.238524086495912e-05, "loss": 0.0258, "step": 101420 }, { "epoch": 2.0268974061788096, "grad_norm": 0.11992501467466354, "learning_rate": 3.237556959056518e-05, "loss": 0.0243, "step": 101430 }, { "epoch": 2.0270972383198114, "grad_norm": 0.12875138223171234, "learning_rate": 3.236589906905453e-05, "loss": 0.0203, "step": 101440 }, { "epoch": 2.027297070460813, "grad_norm": 0.12421323359012604, "learning_rate": 3.235622930084026e-05, "loss": 0.0233, "step": 101450 }, { "epoch": 2.0274969026018144, "grad_norm": 0.14321859180927277, "learning_rate": 3.234656028633549e-05, "loss": 0.0234, "step": 101460 }, { "epoch": 2.027696734742816, "grad_norm": 0.11917857080698013, "learning_rate": 3.2336892025953215e-05, "loss": 0.0307, "step": 101470 }, { "epoch": 2.027896566883818, "grad_norm": 0.1321994513273239, "learning_rate": 3.2327224520106477e-05, "loss": 0.0363, "step": 101480 }, { "epoch": 2.028096399024819, "grad_norm": 0.11055658012628555, "learning_rate": 3.231755776920823e-05, "loss": 0.0267, "step": 101490 }, { "epoch": 2.028296231165821, "grad_norm": 0.15064115822315216, "learning_rate": 3.230789177367143e-05, "loss": 0.0271, "step": 101500 }, { "epoch": 2.028496063306822, "grad_norm": 0.11690524965524673, "learning_rate": 3.2298226533909e-05, "loss": 0.0295, "step": 101510 }, { "epoch": 2.028695895447824, "grad_norm": 0.11218038946390152, "learning_rate": 3.228856205033381e-05, "loss": 0.0199, "step": 101520 }, { "epoch": 2.0288957275888255, "grad_norm": 0.10350265353918076, "learning_rate": 3.227889832335872e-05, "loss": 0.0231, "step": 101530 }, { "epoch": 2.029095559729827, "grad_norm": 0.08993250131607056, "learning_rate": 3.226923535339652e-05, "loss": 0.0253, "step": 101540 }, { "epoch": 2.0292953918708285, "grad_norm": 0.10517027229070663, "learning_rate": 3.2259573140860025e-05, "loss": 0.0308, "step": 101550 }, { "epoch": 2.0294952240118302, "grad_norm": 0.1455715149641037, "learning_rate": 3.2249911686161984e-05, "loss": 0.0217, "step": 101560 }, { "epoch": 2.0296950561528315, "grad_norm": 0.11619777232408524, "learning_rate": 3.224025098971509e-05, "loss": 0.0247, "step": 101570 }, { "epoch": 2.0298948882938332, "grad_norm": 0.1110013872385025, "learning_rate": 3.223059105193207e-05, "loss": 0.0194, "step": 101580 }, { "epoch": 2.0300947204348345, "grad_norm": 0.11407394707202911, "learning_rate": 3.222093187322554e-05, "loss": 0.0284, "step": 101590 }, { "epoch": 2.0302945525758362, "grad_norm": 0.11705561727285385, "learning_rate": 3.221127345400813e-05, "loss": 0.0215, "step": 101600 }, { "epoch": 2.030494384716838, "grad_norm": 0.15642492473125458, "learning_rate": 3.2201615794692465e-05, "loss": 0.0254, "step": 101610 }, { "epoch": 2.0306942168578392, "grad_norm": 0.14236611127853394, "learning_rate": 3.219195889569106e-05, "loss": 0.0383, "step": 101620 }, { "epoch": 2.030894048998841, "grad_norm": 0.14471547305583954, "learning_rate": 3.218230275741645e-05, "loss": 0.0254, "step": 101630 }, { "epoch": 2.0310938811398427, "grad_norm": 0.13176307082176208, "learning_rate": 3.2172647380281154e-05, "loss": 0.029, "step": 101640 }, { "epoch": 2.031293713280844, "grad_norm": 0.10687657445669174, "learning_rate": 3.21629927646976e-05, "loss": 0.0333, "step": 101650 }, { "epoch": 2.0314935454218457, "grad_norm": 0.17570911347866058, "learning_rate": 3.215333891107823e-05, "loss": 0.0283, "step": 101660 }, { "epoch": 2.0316933775628474, "grad_norm": 0.1416112333536148, "learning_rate": 3.214368581983544e-05, "loss": 0.0313, "step": 101670 }, { "epoch": 2.0318932097038487, "grad_norm": 0.11240031570196152, "learning_rate": 3.213403349138158e-05, "loss": 0.0293, "step": 101680 }, { "epoch": 2.0320930418448504, "grad_norm": 0.13934865593910217, "learning_rate": 3.2124381926128994e-05, "loss": 0.0284, "step": 101690 }, { "epoch": 2.0322928739858517, "grad_norm": 0.177777498960495, "learning_rate": 3.2114731124489975e-05, "loss": 0.0358, "step": 101700 }, { "epoch": 2.0324927061268534, "grad_norm": 0.16926313936710358, "learning_rate": 3.210508108687679e-05, "loss": 0.0219, "step": 101710 }, { "epoch": 2.032692538267855, "grad_norm": 0.1221354603767395, "learning_rate": 3.2095431813701634e-05, "loss": 0.02, "step": 101720 }, { "epoch": 2.0328923704088564, "grad_norm": 0.09624738991260529, "learning_rate": 3.2085783305376774e-05, "loss": 0.0223, "step": 101730 }, { "epoch": 2.033092202549858, "grad_norm": 0.09632746875286102, "learning_rate": 3.207613556231433e-05, "loss": 0.0223, "step": 101740 }, { "epoch": 2.03329203469086, "grad_norm": 0.1433020383119583, "learning_rate": 3.206648858492647e-05, "loss": 0.0236, "step": 101750 }, { "epoch": 2.033491866831861, "grad_norm": 0.12891079485416412, "learning_rate": 3.205684237362524e-05, "loss": 0.0275, "step": 101760 }, { "epoch": 2.033691698972863, "grad_norm": 0.09763986617326736, "learning_rate": 3.2047196928822756e-05, "loss": 0.0348, "step": 101770 }, { "epoch": 2.0338915311138646, "grad_norm": 0.1274484246969223, "learning_rate": 3.203755225093104e-05, "loss": 0.0296, "step": 101780 }, { "epoch": 2.034091363254866, "grad_norm": 0.1153603196144104, "learning_rate": 3.202790834036209e-05, "loss": 0.0291, "step": 101790 }, { "epoch": 2.0342911953958676, "grad_norm": 0.1088789850473404, "learning_rate": 3.201826519752789e-05, "loss": 0.0258, "step": 101800 }, { "epoch": 2.034491027536869, "grad_norm": 0.11084318161010742, "learning_rate": 3.200862282284035e-05, "loss": 0.0239, "step": 101810 }, { "epoch": 2.0346908596778706, "grad_norm": 0.13770125806331635, "learning_rate": 3.19989812167114e-05, "loss": 0.0242, "step": 101820 }, { "epoch": 2.0348906918188723, "grad_norm": 0.1434858739376068, "learning_rate": 3.1989340379552914e-05, "loss": 0.0247, "step": 101830 }, { "epoch": 2.0350905239598736, "grad_norm": 0.09395375847816467, "learning_rate": 3.197970031177671e-05, "loss": 0.0212, "step": 101840 }, { "epoch": 2.0352903561008753, "grad_norm": 0.15505559742450714, "learning_rate": 3.197006101379458e-05, "loss": 0.027, "step": 101850 }, { "epoch": 2.035490188241877, "grad_norm": 0.17794016003608704, "learning_rate": 3.196042248601836e-05, "loss": 0.0382, "step": 101860 }, { "epoch": 2.0356900203828783, "grad_norm": 0.1738266944885254, "learning_rate": 3.195078472885974e-05, "loss": 0.0213, "step": 101870 }, { "epoch": 2.03588985252388, "grad_norm": 0.10415098816156387, "learning_rate": 3.194114774273044e-05, "loss": 0.0285, "step": 101880 }, { "epoch": 2.0360896846648817, "grad_norm": 0.08642081171274185, "learning_rate": 3.1931511528042136e-05, "loss": 0.031, "step": 101890 }, { "epoch": 2.036289516805883, "grad_norm": 0.1303432136774063, "learning_rate": 3.192187608520645e-05, "loss": 0.0282, "step": 101900 }, { "epoch": 2.0364893489468847, "grad_norm": 0.14997880160808563, "learning_rate": 3.191224141463502e-05, "loss": 0.0273, "step": 101910 }, { "epoch": 2.036689181087886, "grad_norm": 0.12313885241746902, "learning_rate": 3.19026075167394e-05, "loss": 0.0312, "step": 101920 }, { "epoch": 2.0368890132288877, "grad_norm": 0.137616366147995, "learning_rate": 3.189297439193116e-05, "loss": 0.0251, "step": 101930 }, { "epoch": 2.0370888453698894, "grad_norm": 0.12340866029262543, "learning_rate": 3.1883342040621764e-05, "loss": 0.0315, "step": 101940 }, { "epoch": 2.0372886775108907, "grad_norm": 0.12556247413158417, "learning_rate": 3.187371046322273e-05, "loss": 0.0277, "step": 101950 }, { "epoch": 2.0374885096518924, "grad_norm": 0.10983385890722275, "learning_rate": 3.186407966014549e-05, "loss": 0.0256, "step": 101960 }, { "epoch": 2.037688341792894, "grad_norm": 0.11596115678548813, "learning_rate": 3.185444963180144e-05, "loss": 0.0336, "step": 101970 }, { "epoch": 2.0378881739338954, "grad_norm": 0.12201960384845734, "learning_rate": 3.184482037860198e-05, "loss": 0.0315, "step": 101980 }, { "epoch": 2.038088006074897, "grad_norm": 0.10136423259973526, "learning_rate": 3.183519190095843e-05, "loss": 0.0347, "step": 101990 }, { "epoch": 2.0382878382158984, "grad_norm": 0.10785144567489624, "learning_rate": 3.182556419928211e-05, "loss": 0.0282, "step": 102000 }, { "epoch": 2.0384876703569, "grad_norm": 0.10069535672664642, "learning_rate": 3.181593727398431e-05, "loss": 0.0195, "step": 102010 }, { "epoch": 2.038687502497902, "grad_norm": 0.13615106046199799, "learning_rate": 3.180631112547626e-05, "loss": 0.0189, "step": 102020 }, { "epoch": 2.038887334638903, "grad_norm": 0.16936446726322174, "learning_rate": 3.179668575416917e-05, "loss": 0.0338, "step": 102030 }, { "epoch": 2.039087166779905, "grad_norm": 0.1202886626124382, "learning_rate": 3.1787061160474226e-05, "loss": 0.0271, "step": 102040 }, { "epoch": 2.0392869989209066, "grad_norm": 0.09991130232810974, "learning_rate": 3.1777437344802576e-05, "loss": 0.0248, "step": 102050 }, { "epoch": 2.039486831061908, "grad_norm": 0.13650737702846527, "learning_rate": 3.1767814307565337e-05, "loss": 0.0354, "step": 102060 }, { "epoch": 2.0396866632029096, "grad_norm": 0.11697321385145187, "learning_rate": 3.1758192049173564e-05, "loss": 0.0255, "step": 102070 }, { "epoch": 2.0398864953439113, "grad_norm": 0.13735759258270264, "learning_rate": 3.174857057003831e-05, "loss": 0.0275, "step": 102080 }, { "epoch": 2.0400863274849126, "grad_norm": 0.11231206357479095, "learning_rate": 3.1738949870570586e-05, "loss": 0.0315, "step": 102090 }, { "epoch": 2.0402861596259143, "grad_norm": 0.10042896121740341, "learning_rate": 3.17293299511814e-05, "loss": 0.021, "step": 102100 }, { "epoch": 2.0404859917669156, "grad_norm": 0.12027310580015182, "learning_rate": 3.171971081228167e-05, "loss": 0.0275, "step": 102110 }, { "epoch": 2.0406858239079173, "grad_norm": 0.12280505150556564, "learning_rate": 3.171009245428229e-05, "loss": 0.0289, "step": 102120 }, { "epoch": 2.040885656048919, "grad_norm": 0.13032837212085724, "learning_rate": 3.170047487759418e-05, "loss": 0.0305, "step": 102130 }, { "epoch": 2.0410854881899203, "grad_norm": 0.14005929231643677, "learning_rate": 3.169085808262816e-05, "loss": 0.0275, "step": 102140 }, { "epoch": 2.041285320330922, "grad_norm": 0.08877997100353241, "learning_rate": 3.168124206979507e-05, "loss": 0.0211, "step": 102150 }, { "epoch": 2.0414851524719237, "grad_norm": 0.11179397255182266, "learning_rate": 3.1671626839505645e-05, "loss": 0.033, "step": 102160 }, { "epoch": 2.041684984612925, "grad_norm": 0.10337316244840622, "learning_rate": 3.166201239217064e-05, "loss": 0.0315, "step": 102170 }, { "epoch": 2.0418848167539267, "grad_norm": 0.1856849193572998, "learning_rate": 3.1652398728200794e-05, "loss": 0.0289, "step": 102180 }, { "epoch": 2.0420846488949285, "grad_norm": 0.11861107498407364, "learning_rate": 3.1642785848006764e-05, "loss": 0.0212, "step": 102190 }, { "epoch": 2.0422844810359297, "grad_norm": 0.12747997045516968, "learning_rate": 3.1633173751999205e-05, "loss": 0.0273, "step": 102200 }, { "epoch": 2.0424843131769315, "grad_norm": 0.17700806260108948, "learning_rate": 3.16235624405887e-05, "loss": 0.0343, "step": 102210 }, { "epoch": 2.0426841453179327, "grad_norm": 0.15275658667087555, "learning_rate": 3.161395191418587e-05, "loss": 0.0301, "step": 102220 }, { "epoch": 2.0428839774589345, "grad_norm": 0.10694307088851929, "learning_rate": 3.160434217320124e-05, "loss": 0.0243, "step": 102230 }, { "epoch": 2.043083809599936, "grad_norm": 0.11009056866168976, "learning_rate": 3.1594733218045316e-05, "loss": 0.0224, "step": 102240 }, { "epoch": 2.0432836417409375, "grad_norm": 0.10862893611192703, "learning_rate": 3.158512504912858e-05, "loss": 0.0283, "step": 102250 }, { "epoch": 2.043483473881939, "grad_norm": 0.1514536738395691, "learning_rate": 3.157551766686146e-05, "loss": 0.0348, "step": 102260 }, { "epoch": 2.043683306022941, "grad_norm": 0.10862571746110916, "learning_rate": 3.1565911071654385e-05, "loss": 0.028, "step": 102270 }, { "epoch": 2.043883138163942, "grad_norm": 0.1229274645447731, "learning_rate": 3.155630526391773e-05, "loss": 0.0252, "step": 102280 }, { "epoch": 2.044082970304944, "grad_norm": 0.11733239889144897, "learning_rate": 3.1546700244061824e-05, "loss": 0.0211, "step": 102290 }, { "epoch": 2.0442828024459456, "grad_norm": 0.09758755564689636, "learning_rate": 3.153709601249698e-05, "loss": 0.0276, "step": 102300 }, { "epoch": 2.044482634586947, "grad_norm": 0.13322186470031738, "learning_rate": 3.152749256963349e-05, "loss": 0.022, "step": 102310 }, { "epoch": 2.0446824667279486, "grad_norm": 0.09929981082677841, "learning_rate": 3.1517889915881576e-05, "loss": 0.0341, "step": 102320 }, { "epoch": 2.04488229886895, "grad_norm": 0.1041535809636116, "learning_rate": 3.150828805165146e-05, "loss": 0.0351, "step": 102330 }, { "epoch": 2.0450821310099516, "grad_norm": 0.09500256925821304, "learning_rate": 3.1498686977353295e-05, "loss": 0.0287, "step": 102340 }, { "epoch": 2.0452819631509533, "grad_norm": 0.09905274212360382, "learning_rate": 3.148908669339723e-05, "loss": 0.0295, "step": 102350 }, { "epoch": 2.0454817952919546, "grad_norm": 0.11479157209396362, "learning_rate": 3.147948720019339e-05, "loss": 0.0271, "step": 102360 }, { "epoch": 2.0456816274329563, "grad_norm": 0.17361022531986237, "learning_rate": 3.1469888498151835e-05, "loss": 0.0387, "step": 102370 }, { "epoch": 2.045881459573958, "grad_norm": 0.10727051645517349, "learning_rate": 3.1460290587682606e-05, "loss": 0.0199, "step": 102380 }, { "epoch": 2.0460812917149593, "grad_norm": 0.1571158766746521, "learning_rate": 3.145069346919568e-05, "loss": 0.031, "step": 102390 }, { "epoch": 2.046281123855961, "grad_norm": 0.12198363244533539, "learning_rate": 3.1441097143101064e-05, "loss": 0.0236, "step": 102400 }, { "epoch": 2.0464809559969623, "grad_norm": 0.12463497370481491, "learning_rate": 3.14315016098087e-05, "loss": 0.0218, "step": 102410 }, { "epoch": 2.046680788137964, "grad_norm": 0.10220472514629364, "learning_rate": 3.142190686972846e-05, "loss": 0.0254, "step": 102420 }, { "epoch": 2.046880620278966, "grad_norm": 0.13399823009967804, "learning_rate": 3.1412312923270246e-05, "loss": 0.0355, "step": 102430 }, { "epoch": 2.047080452419967, "grad_norm": 0.09064894169569016, "learning_rate": 3.140271977084385e-05, "loss": 0.0227, "step": 102440 }, { "epoch": 2.0472802845609688, "grad_norm": 0.10963297635316849, "learning_rate": 3.1393127412859115e-05, "loss": 0.0281, "step": 102450 }, { "epoch": 2.0474801167019705, "grad_norm": 0.13876476883888245, "learning_rate": 3.13835358497258e-05, "loss": 0.0267, "step": 102460 }, { "epoch": 2.0476799488429718, "grad_norm": 0.11889202147722244, "learning_rate": 3.137394508185363e-05, "loss": 0.0244, "step": 102470 }, { "epoch": 2.0478797809839735, "grad_norm": 0.12479867041110992, "learning_rate": 3.13643551096523e-05, "loss": 0.0248, "step": 102480 }, { "epoch": 2.048079613124975, "grad_norm": 0.13938482105731964, "learning_rate": 3.135476593353149e-05, "loss": 0.0277, "step": 102490 }, { "epoch": 2.0482794452659765, "grad_norm": 0.09752371907234192, "learning_rate": 3.134517755390084e-05, "loss": 0.0229, "step": 102500 }, { "epoch": 2.048479277406978, "grad_norm": 0.16607587039470673, "learning_rate": 3.1335589971169924e-05, "loss": 0.0269, "step": 102510 }, { "epoch": 2.0486791095479795, "grad_norm": 0.12218862771987915, "learning_rate": 3.132600318574832e-05, "loss": 0.0293, "step": 102520 }, { "epoch": 2.048878941688981, "grad_norm": 0.11881019175052643, "learning_rate": 3.131641719804554e-05, "loss": 0.0213, "step": 102530 }, { "epoch": 2.049078773829983, "grad_norm": 0.15054476261138916, "learning_rate": 3.130683200847111e-05, "loss": 0.0264, "step": 102540 }, { "epoch": 2.049278605970984, "grad_norm": 0.11551321297883987, "learning_rate": 3.129724761743448e-05, "loss": 0.0245, "step": 102550 }, { "epoch": 2.049478438111986, "grad_norm": 0.12811435759067535, "learning_rate": 3.1287664025345066e-05, "loss": 0.0269, "step": 102560 }, { "epoch": 2.0496782702529877, "grad_norm": 0.16151680052280426, "learning_rate": 3.127808123261226e-05, "loss": 0.0268, "step": 102570 }, { "epoch": 2.049878102393989, "grad_norm": 0.1493188589811325, "learning_rate": 3.1268499239645446e-05, "loss": 0.023, "step": 102580 }, { "epoch": 2.0500779345349907, "grad_norm": 0.1533786803483963, "learning_rate": 3.125891804685393e-05, "loss": 0.0254, "step": 102590 }, { "epoch": 2.0502777666759924, "grad_norm": 0.11143757402896881, "learning_rate": 3.124933765464702e-05, "loss": 0.031, "step": 102600 }, { "epoch": 2.0504775988169937, "grad_norm": 0.12095695734024048, "learning_rate": 3.1239758063433935e-05, "loss": 0.024, "step": 102610 }, { "epoch": 2.0506774309579954, "grad_norm": 0.10568098723888397, "learning_rate": 3.1230179273623925e-05, "loss": 0.0238, "step": 102620 }, { "epoch": 2.0508772630989967, "grad_norm": 0.1318988800048828, "learning_rate": 3.1220601285626184e-05, "loss": 0.023, "step": 102630 }, { "epoch": 2.0510770952399984, "grad_norm": 0.114771269261837, "learning_rate": 3.121102409984986e-05, "loss": 0.0259, "step": 102640 }, { "epoch": 2.051276927381, "grad_norm": 0.15181897580623627, "learning_rate": 3.120144771670407e-05, "loss": 0.027, "step": 102650 }, { "epoch": 2.0514767595220014, "grad_norm": 0.13417527079582214, "learning_rate": 3.119187213659788e-05, "loss": 0.026, "step": 102660 }, { "epoch": 2.051676591663003, "grad_norm": 0.10252025723457336, "learning_rate": 3.118229735994037e-05, "loss": 0.024, "step": 102670 }, { "epoch": 2.051876423804005, "grad_norm": 0.1330791860818863, "learning_rate": 3.1172723387140566e-05, "loss": 0.0328, "step": 102680 }, { "epoch": 2.052076255945006, "grad_norm": 0.11593128740787506, "learning_rate": 3.1163150218607416e-05, "loss": 0.0269, "step": 102690 }, { "epoch": 2.052276088086008, "grad_norm": 0.10161211341619492, "learning_rate": 3.1153577854749895e-05, "loss": 0.0197, "step": 102700 }, { "epoch": 2.0524759202270095, "grad_norm": 0.09545958042144775, "learning_rate": 3.114400629597689e-05, "loss": 0.0267, "step": 102710 }, { "epoch": 2.052675752368011, "grad_norm": 0.17067717015743256, "learning_rate": 3.1134435542697296e-05, "loss": 0.0363, "step": 102720 }, { "epoch": 2.0528755845090125, "grad_norm": 0.14066657423973083, "learning_rate": 3.112486559531997e-05, "loss": 0.0281, "step": 102730 }, { "epoch": 2.053075416650014, "grad_norm": 0.13771016895771027, "learning_rate": 3.111529645425371e-05, "loss": 0.0284, "step": 102740 }, { "epoch": 2.0532752487910155, "grad_norm": 0.1398041546344757, "learning_rate": 3.1105728119907284e-05, "loss": 0.0246, "step": 102750 }, { "epoch": 2.0534750809320172, "grad_norm": 0.44648686051368713, "learning_rate": 3.109616059268946e-05, "loss": 0.0291, "step": 102760 }, { "epoch": 2.0536749130730185, "grad_norm": 0.14801038801670074, "learning_rate": 3.108659387300891e-05, "loss": 0.0283, "step": 102770 }, { "epoch": 2.0538747452140202, "grad_norm": 0.11700168997049332, "learning_rate": 3.107702796127434e-05, "loss": 0.0312, "step": 102780 }, { "epoch": 2.054074577355022, "grad_norm": 0.14393997192382812, "learning_rate": 3.106746285789437e-05, "loss": 0.0268, "step": 102790 }, { "epoch": 2.0542744094960232, "grad_norm": 0.1321331113576889, "learning_rate": 3.105789856327759e-05, "loss": 0.0248, "step": 102800 }, { "epoch": 2.054474241637025, "grad_norm": 0.19026395678520203, "learning_rate": 3.1048335077832605e-05, "loss": 0.0286, "step": 102810 }, { "epoch": 2.0546740737780262, "grad_norm": 0.1377466470003128, "learning_rate": 3.1038772401967917e-05, "loss": 0.0246, "step": 102820 }, { "epoch": 2.054873905919028, "grad_norm": 0.11270447075366974, "learning_rate": 3.102921053609206e-05, "loss": 0.033, "step": 102830 }, { "epoch": 2.0550737380600297, "grad_norm": 0.10397548228502274, "learning_rate": 3.1019649480613444e-05, "loss": 0.0272, "step": 102840 }, { "epoch": 2.055273570201031, "grad_norm": 0.10235434770584106, "learning_rate": 3.101008923594055e-05, "loss": 0.0266, "step": 102850 }, { "epoch": 2.0554734023420327, "grad_norm": 0.11493117362260818, "learning_rate": 3.100052980248177e-05, "loss": 0.0219, "step": 102860 }, { "epoch": 2.0556732344830344, "grad_norm": 0.1325855404138565, "learning_rate": 3.0990971180645445e-05, "loss": 0.0313, "step": 102870 }, { "epoch": 2.0558730666240357, "grad_norm": 0.11929338425397873, "learning_rate": 3.0981413370839925e-05, "loss": 0.0231, "step": 102880 }, { "epoch": 2.0560728987650374, "grad_norm": 0.13147883117198944, "learning_rate": 3.097185637347345e-05, "loss": 0.0246, "step": 102890 }, { "epoch": 2.056272730906039, "grad_norm": 0.0986345037817955, "learning_rate": 3.096230018895435e-05, "loss": 0.0247, "step": 102900 }, { "epoch": 2.0564725630470404, "grad_norm": 0.1007135733962059, "learning_rate": 3.095274481769079e-05, "loss": 0.0203, "step": 102910 }, { "epoch": 2.056672395188042, "grad_norm": 0.11158177256584167, "learning_rate": 3.0943190260091006e-05, "loss": 0.0257, "step": 102920 }, { "epoch": 2.0568722273290434, "grad_norm": 0.1355234831571579, "learning_rate": 3.09336365165631e-05, "loss": 0.0259, "step": 102930 }, { "epoch": 2.057072059470045, "grad_norm": 0.10580695420503616, "learning_rate": 3.0924083587515216e-05, "loss": 0.0294, "step": 102940 }, { "epoch": 2.057271891611047, "grad_norm": 0.1114768460392952, "learning_rate": 3.091453147335546e-05, "loss": 0.0205, "step": 102950 }, { "epoch": 2.057471723752048, "grad_norm": 0.14107394218444824, "learning_rate": 3.090498017449185e-05, "loss": 0.0333, "step": 102960 }, { "epoch": 2.05767155589305, "grad_norm": 0.12760888040065765, "learning_rate": 3.089542969133239e-05, "loss": 0.0268, "step": 102970 }, { "epoch": 2.0578713880340516, "grad_norm": 0.11440987884998322, "learning_rate": 3.08858800242851e-05, "loss": 0.0317, "step": 102980 }, { "epoch": 2.058071220175053, "grad_norm": 0.07452709227800369, "learning_rate": 3.087633117375789e-05, "loss": 0.0198, "step": 102990 }, { "epoch": 2.0582710523160546, "grad_norm": 0.11310607939958572, "learning_rate": 3.086678314015869e-05, "loss": 0.0235, "step": 103000 }, { "epoch": 2.0584708844570563, "grad_norm": 0.23484580218791962, "learning_rate": 3.085723592389537e-05, "loss": 0.0228, "step": 103010 }, { "epoch": 2.0586707165980576, "grad_norm": 0.327136367559433, "learning_rate": 3.084768952537574e-05, "loss": 0.0274, "step": 103020 }, { "epoch": 2.0588705487390593, "grad_norm": 0.0965195894241333, "learning_rate": 3.083814394500765e-05, "loss": 0.0231, "step": 103030 }, { "epoch": 2.0590703808800606, "grad_norm": 0.12499932944774628, "learning_rate": 3.082859918319885e-05, "loss": 0.0299, "step": 103040 }, { "epoch": 2.0592702130210623, "grad_norm": 0.11812655627727509, "learning_rate": 3.081905524035707e-05, "loss": 0.025, "step": 103050 }, { "epoch": 2.059470045162064, "grad_norm": 0.11479450017213821, "learning_rate": 3.080951211689001e-05, "loss": 0.029, "step": 103060 }, { "epoch": 2.0596698773030653, "grad_norm": 0.14315302670001984, "learning_rate": 3.079996981320533e-05, "loss": 0.0241, "step": 103070 }, { "epoch": 2.059869709444067, "grad_norm": 0.1164930909872055, "learning_rate": 3.0790428329710684e-05, "loss": 0.0314, "step": 103080 }, { "epoch": 2.0600695415850687, "grad_norm": 0.09979183971881866, "learning_rate": 3.078088766681364e-05, "loss": 0.0249, "step": 103090 }, { "epoch": 2.06026937372607, "grad_norm": 0.11825120449066162, "learning_rate": 3.0771347824921786e-05, "loss": 0.0264, "step": 103100 }, { "epoch": 2.0604692058670717, "grad_norm": 0.10855257511138916, "learning_rate": 3.07618088044426e-05, "loss": 0.0269, "step": 103110 }, { "epoch": 2.060669038008073, "grad_norm": 0.10649462789297104, "learning_rate": 3.075227060578362e-05, "loss": 0.0265, "step": 103120 }, { "epoch": 2.0608688701490747, "grad_norm": 0.1283763200044632, "learning_rate": 3.074273322935228e-05, "loss": 0.0258, "step": 103130 }, { "epoch": 2.0610687022900764, "grad_norm": 0.09683689475059509, "learning_rate": 3.0733196675556e-05, "loss": 0.0318, "step": 103140 }, { "epoch": 2.0612685344310777, "grad_norm": 0.11702023446559906, "learning_rate": 3.072366094480215e-05, "loss": 0.035, "step": 103150 }, { "epoch": 2.0614683665720794, "grad_norm": 0.13139821588993073, "learning_rate": 3.0714126037498104e-05, "loss": 0.0337, "step": 103160 }, { "epoch": 2.061668198713081, "grad_norm": 0.14203286170959473, "learning_rate": 3.070459195405116e-05, "loss": 0.0335, "step": 103170 }, { "epoch": 2.0618680308540824, "grad_norm": 0.10834820568561554, "learning_rate": 3.069505869486862e-05, "loss": 0.024, "step": 103180 }, { "epoch": 2.062067862995084, "grad_norm": 0.10928278416395187, "learning_rate": 3.068552626035769e-05, "loss": 0.027, "step": 103190 }, { "epoch": 2.062267695136086, "grad_norm": 0.11829359829425812, "learning_rate": 3.067599465092558e-05, "loss": 0.0247, "step": 103200 }, { "epoch": 2.062467527277087, "grad_norm": 0.14860375225543976, "learning_rate": 3.066646386697951e-05, "loss": 0.0323, "step": 103210 }, { "epoch": 2.062667359418089, "grad_norm": 0.09753859043121338, "learning_rate": 3.0656933908926564e-05, "loss": 0.0249, "step": 103220 }, { "epoch": 2.06286719155909, "grad_norm": 0.12596248090267181, "learning_rate": 3.0647404777173896e-05, "loss": 0.0175, "step": 103230 }, { "epoch": 2.063067023700092, "grad_norm": 0.09862314909696579, "learning_rate": 3.0637876472128516e-05, "loss": 0.0246, "step": 103240 }, { "epoch": 2.0632668558410936, "grad_norm": 0.10416673868894577, "learning_rate": 3.0628348994197486e-05, "loss": 0.0284, "step": 103250 }, { "epoch": 2.063466687982095, "grad_norm": 0.11005798727273941, "learning_rate": 3.061882234378781e-05, "loss": 0.0206, "step": 103260 }, { "epoch": 2.0636665201230966, "grad_norm": 0.12117450684309006, "learning_rate": 3.060929652130644e-05, "loss": 0.0295, "step": 103270 }, { "epoch": 2.0638663522640983, "grad_norm": 0.17923074960708618, "learning_rate": 3.0599771527160305e-05, "loss": 0.0268, "step": 103280 }, { "epoch": 2.0640661844050996, "grad_norm": 0.11297452449798584, "learning_rate": 3.0590247361756265e-05, "loss": 0.0251, "step": 103290 }, { "epoch": 2.0642660165461013, "grad_norm": 0.09386575222015381, "learning_rate": 3.0580724025501224e-05, "loss": 0.0245, "step": 103300 }, { "epoch": 2.064465848687103, "grad_norm": 0.1671212911605835, "learning_rate": 3.0571201518801974e-05, "loss": 0.0208, "step": 103310 }, { "epoch": 2.0646656808281043, "grad_norm": 0.12314672768115997, "learning_rate": 3.056167984206531e-05, "loss": 0.0255, "step": 103320 }, { "epoch": 2.064865512969106, "grad_norm": 0.13050271570682526, "learning_rate": 3.055215899569796e-05, "loss": 0.0305, "step": 103330 }, { "epoch": 2.0650653451101073, "grad_norm": 0.10591874271631241, "learning_rate": 3.0542638980106646e-05, "loss": 0.0322, "step": 103340 }, { "epoch": 2.065265177251109, "grad_norm": 0.15772776305675507, "learning_rate": 3.0533119795698074e-05, "loss": 0.0292, "step": 103350 }, { "epoch": 2.0654650093921108, "grad_norm": 0.11265631020069122, "learning_rate": 3.0523601442878843e-05, "loss": 0.0304, "step": 103360 }, { "epoch": 2.065664841533112, "grad_norm": 0.1107310876250267, "learning_rate": 3.0514083922055597e-05, "loss": 0.0217, "step": 103370 }, { "epoch": 2.0658646736741137, "grad_norm": 0.11067664623260498, "learning_rate": 3.0504567233634867e-05, "loss": 0.0284, "step": 103380 }, { "epoch": 2.0660645058151155, "grad_norm": 0.1874716430902481, "learning_rate": 3.049505137802322e-05, "loss": 0.0325, "step": 103390 }, { "epoch": 2.0662643379561167, "grad_norm": 0.12870238721370697, "learning_rate": 3.0485536355627152e-05, "loss": 0.0211, "step": 103400 }, { "epoch": 2.0664641700971185, "grad_norm": 0.0960896760225296, "learning_rate": 3.0476022166853125e-05, "loss": 0.0247, "step": 103410 }, { "epoch": 2.06666400223812, "grad_norm": 0.131771519780159, "learning_rate": 3.046650881210754e-05, "loss": 0.0244, "step": 103420 }, { "epoch": 2.0668638343791215, "grad_norm": 0.12263655662536621, "learning_rate": 3.0456996291796842e-05, "loss": 0.0268, "step": 103430 }, { "epoch": 2.067063666520123, "grad_norm": 0.12969480454921722, "learning_rate": 3.0447484606327348e-05, "loss": 0.029, "step": 103440 }, { "epoch": 2.0672634986611245, "grad_norm": 0.12007904052734375, "learning_rate": 3.0437973756105402e-05, "loss": 0.0238, "step": 103450 }, { "epoch": 2.067463330802126, "grad_norm": 0.10252724587917328, "learning_rate": 3.042846374153727e-05, "loss": 0.0215, "step": 103460 }, { "epoch": 2.067663162943128, "grad_norm": 0.10874901711940765, "learning_rate": 3.0418954563029213e-05, "loss": 0.0258, "step": 103470 }, { "epoch": 2.067862995084129, "grad_norm": 0.13154760003089905, "learning_rate": 3.0409446220987453e-05, "loss": 0.0403, "step": 103480 }, { "epoch": 2.068062827225131, "grad_norm": 0.10961780697107315, "learning_rate": 3.0399938715818156e-05, "loss": 0.0257, "step": 103490 }, { "epoch": 2.0682626593661326, "grad_norm": 0.14498132467269897, "learning_rate": 3.039043204792748e-05, "loss": 0.0237, "step": 103500 }, { "epoch": 2.068462491507134, "grad_norm": 0.1280423253774643, "learning_rate": 3.0380926217721505e-05, "loss": 0.028, "step": 103510 }, { "epoch": 2.0686623236481356, "grad_norm": 0.14965607225894928, "learning_rate": 3.0371421225606324e-05, "loss": 0.0255, "step": 103520 }, { "epoch": 2.0688621557891373, "grad_norm": 0.1479950249195099, "learning_rate": 3.036191707198799e-05, "loss": 0.0277, "step": 103530 }, { "epoch": 2.0690619879301386, "grad_norm": 0.11392126977443695, "learning_rate": 3.0352413757272463e-05, "loss": 0.024, "step": 103540 }, { "epoch": 2.0692618200711403, "grad_norm": 0.10055316239595413, "learning_rate": 3.0342911281865738e-05, "loss": 0.0312, "step": 103550 }, { "epoch": 2.0694616522121416, "grad_norm": 0.10480648279190063, "learning_rate": 3.0333409646173715e-05, "loss": 0.0282, "step": 103560 }, { "epoch": 2.0696614843531433, "grad_norm": 0.11559252440929413, "learning_rate": 3.032390885060231e-05, "loss": 0.0271, "step": 103570 }, { "epoch": 2.069861316494145, "grad_norm": 0.14059817790985107, "learning_rate": 3.031440889555739e-05, "loss": 0.026, "step": 103580 }, { "epoch": 2.0700611486351463, "grad_norm": 0.12318948656320572, "learning_rate": 3.030490978144474e-05, "loss": 0.0315, "step": 103590 }, { "epoch": 2.070260980776148, "grad_norm": 0.11673494428396225, "learning_rate": 3.0295411508670158e-05, "loss": 0.0299, "step": 103600 }, { "epoch": 2.07046081291715, "grad_norm": 0.12787385284900665, "learning_rate": 3.0285914077639415e-05, "loss": 0.0258, "step": 103610 }, { "epoch": 2.070660645058151, "grad_norm": 0.206668421626091, "learning_rate": 3.02764174887582e-05, "loss": 0.0202, "step": 103620 }, { "epoch": 2.070860477199153, "grad_norm": 0.19779261946678162, "learning_rate": 3.0266921742432202e-05, "loss": 0.0293, "step": 103630 }, { "epoch": 2.071060309340154, "grad_norm": 0.09301337599754333, "learning_rate": 3.025742683906705e-05, "loss": 0.0255, "step": 103640 }, { "epoch": 2.071260141481156, "grad_norm": 0.17799663543701172, "learning_rate": 3.0247932779068345e-05, "loss": 0.0208, "step": 103650 }, { "epoch": 2.0714599736221575, "grad_norm": 0.08414336293935776, "learning_rate": 3.023843956284168e-05, "loss": 0.0197, "step": 103660 }, { "epoch": 2.071659805763159, "grad_norm": 0.09615986049175262, "learning_rate": 3.0228947190792572e-05, "loss": 0.0291, "step": 103670 }, { "epoch": 2.0718596379041605, "grad_norm": 0.13791023194789886, "learning_rate": 3.0219455663326525e-05, "loss": 0.0269, "step": 103680 }, { "epoch": 2.072059470045162, "grad_norm": 0.09929884225130081, "learning_rate": 3.020996498084897e-05, "loss": 0.0255, "step": 103690 }, { "epoch": 2.0722593021861635, "grad_norm": 0.09986452013254166, "learning_rate": 3.020047514376538e-05, "loss": 0.029, "step": 103700 }, { "epoch": 2.072459134327165, "grad_norm": 0.09558796882629395, "learning_rate": 3.0190986152481104e-05, "loss": 0.0259, "step": 103710 }, { "epoch": 2.072658966468167, "grad_norm": 0.14649702608585358, "learning_rate": 3.0181498007401533e-05, "loss": 0.0237, "step": 103720 }, { "epoch": 2.072858798609168, "grad_norm": 0.12668105959892273, "learning_rate": 3.0172010708931936e-05, "loss": 0.0281, "step": 103730 }, { "epoch": 2.07305863075017, "grad_norm": 0.09076426178216934, "learning_rate": 3.016252425747761e-05, "loss": 0.0265, "step": 103740 }, { "epoch": 2.073258462891171, "grad_norm": 0.1465519815683365, "learning_rate": 3.0153038653443833e-05, "loss": 0.023, "step": 103750 }, { "epoch": 2.073458295032173, "grad_norm": 0.10046517103910446, "learning_rate": 3.0143553897235765e-05, "loss": 0.0206, "step": 103760 }, { "epoch": 2.0736581271731747, "grad_norm": 0.12085262686014175, "learning_rate": 3.0134069989258607e-05, "loss": 0.0313, "step": 103770 }, { "epoch": 2.073857959314176, "grad_norm": 0.09574572741985321, "learning_rate": 3.012458692991747e-05, "loss": 0.0223, "step": 103780 }, { "epoch": 2.0740577914551777, "grad_norm": 0.12311507016420364, "learning_rate": 3.0115104719617477e-05, "loss": 0.0265, "step": 103790 }, { "epoch": 2.0742576235961794, "grad_norm": 0.13184885680675507, "learning_rate": 3.010562335876369e-05, "loss": 0.0232, "step": 103800 }, { "epoch": 2.0744574557371807, "grad_norm": 0.15104325115680695, "learning_rate": 3.009614284776111e-05, "loss": 0.0313, "step": 103810 }, { "epoch": 2.0746572878781824, "grad_norm": 0.13167819380760193, "learning_rate": 3.0086663187014758e-05, "loss": 0.0189, "step": 103820 }, { "epoch": 2.074857120019184, "grad_norm": 0.13313771784305573, "learning_rate": 3.007718437692955e-05, "loss": 0.0207, "step": 103830 }, { "epoch": 2.0750569521601854, "grad_norm": 0.1307484358549118, "learning_rate": 3.006770641791044e-05, "loss": 0.0293, "step": 103840 }, { "epoch": 2.075256784301187, "grad_norm": 0.13156132400035858, "learning_rate": 3.0058229310362307e-05, "loss": 0.0229, "step": 103850 }, { "epoch": 2.0754566164421884, "grad_norm": 0.12742576003074646, "learning_rate": 3.0048753054689964e-05, "loss": 0.0333, "step": 103860 }, { "epoch": 2.07565644858319, "grad_norm": 0.10799320787191391, "learning_rate": 3.0039277651298235e-05, "loss": 0.028, "step": 103870 }, { "epoch": 2.075856280724192, "grad_norm": 0.12769193947315216, "learning_rate": 3.0029803100591915e-05, "loss": 0.021, "step": 103880 }, { "epoch": 2.076056112865193, "grad_norm": 0.15239761769771576, "learning_rate": 3.0020329402975712e-05, "loss": 0.0302, "step": 103890 }, { "epoch": 2.076255945006195, "grad_norm": 0.1403893083333969, "learning_rate": 3.0010856558854338e-05, "loss": 0.0312, "step": 103900 }, { "epoch": 2.0764557771471965, "grad_norm": 0.09041323512792587, "learning_rate": 3.0001384568632452e-05, "loss": 0.0346, "step": 103910 }, { "epoch": 2.076655609288198, "grad_norm": 0.11879768967628479, "learning_rate": 2.9991913432714657e-05, "loss": 0.0234, "step": 103920 }, { "epoch": 2.0768554414291995, "grad_norm": 0.09763437509536743, "learning_rate": 2.998244315150559e-05, "loss": 0.0315, "step": 103930 }, { "epoch": 2.077055273570201, "grad_norm": 0.13052015006542206, "learning_rate": 2.9972973725409765e-05, "loss": 0.0297, "step": 103940 }, { "epoch": 2.0772551057112025, "grad_norm": 0.13692134618759155, "learning_rate": 2.9963505154831716e-05, "loss": 0.0309, "step": 103950 }, { "epoch": 2.0774549378522043, "grad_norm": 0.10704809427261353, "learning_rate": 2.9954037440175907e-05, "loss": 0.0225, "step": 103960 }, { "epoch": 2.0776547699932055, "grad_norm": 0.16143734753131866, "learning_rate": 2.9944570581846802e-05, "loss": 0.0221, "step": 103970 }, { "epoch": 2.0778546021342073, "grad_norm": 0.09184572100639343, "learning_rate": 2.9935104580248806e-05, "loss": 0.0261, "step": 103980 }, { "epoch": 2.078054434275209, "grad_norm": 0.16682207584381104, "learning_rate": 2.9925639435786278e-05, "loss": 0.0224, "step": 103990 }, { "epoch": 2.0782542664162102, "grad_norm": 0.10195364058017731, "learning_rate": 2.9916175148863562e-05, "loss": 0.0304, "step": 104000 }, { "epoch": 2.078454098557212, "grad_norm": 0.11786826699972153, "learning_rate": 2.990671171988493e-05, "loss": 0.0241, "step": 104010 }, { "epoch": 2.0786539306982137, "grad_norm": 0.11086088418960571, "learning_rate": 2.989724914925467e-05, "loss": 0.0249, "step": 104020 }, { "epoch": 2.078853762839215, "grad_norm": 0.11168328672647476, "learning_rate": 2.9887787437377002e-05, "loss": 0.0345, "step": 104030 }, { "epoch": 2.0790535949802167, "grad_norm": 0.13166388869285583, "learning_rate": 2.9878326584656102e-05, "loss": 0.0318, "step": 104040 }, { "epoch": 2.079253427121218, "grad_norm": 0.1406029611825943, "learning_rate": 2.9868866591496116e-05, "loss": 0.0265, "step": 104050 }, { "epoch": 2.0794532592622197, "grad_norm": 0.1306678205728531, "learning_rate": 2.985940745830118e-05, "loss": 0.0291, "step": 104060 }, { "epoch": 2.0796530914032214, "grad_norm": 0.15756405889987946, "learning_rate": 2.9849949185475358e-05, "loss": 0.0246, "step": 104070 }, { "epoch": 2.0798529235442227, "grad_norm": 0.17937758564949036, "learning_rate": 2.984049177342269e-05, "loss": 0.0273, "step": 104080 }, { "epoch": 2.0800527556852244, "grad_norm": 0.13743792474269867, "learning_rate": 2.9831035222547155e-05, "loss": 0.0264, "step": 104090 }, { "epoch": 2.080252587826226, "grad_norm": 0.13012976944446564, "learning_rate": 2.9821579533252775e-05, "loss": 0.0351, "step": 104100 }, { "epoch": 2.0804524199672274, "grad_norm": 0.10559211671352386, "learning_rate": 2.9812124705943433e-05, "loss": 0.0203, "step": 104110 }, { "epoch": 2.080652252108229, "grad_norm": 0.11133861541748047, "learning_rate": 2.9802670741023053e-05, "loss": 0.0198, "step": 104120 }, { "epoch": 2.080852084249231, "grad_norm": 0.09628912806510925, "learning_rate": 2.979321763889546e-05, "loss": 0.029, "step": 104130 }, { "epoch": 2.081051916390232, "grad_norm": 0.10100444406270981, "learning_rate": 2.978376539996448e-05, "loss": 0.0301, "step": 104140 }, { "epoch": 2.081251748531234, "grad_norm": 0.10890176892280579, "learning_rate": 2.9774314024633926e-05, "loss": 0.0186, "step": 104150 }, { "epoch": 2.081451580672235, "grad_norm": 0.1123933494091034, "learning_rate": 2.976486351330751e-05, "loss": 0.0349, "step": 104160 }, { "epoch": 2.081651412813237, "grad_norm": 0.21305227279663086, "learning_rate": 2.975541386638896e-05, "loss": 0.0245, "step": 104170 }, { "epoch": 2.0818512449542386, "grad_norm": 0.11282804608345032, "learning_rate": 2.9745965084281923e-05, "loss": 0.0243, "step": 104180 }, { "epoch": 2.08205107709524, "grad_norm": 0.15626512467861176, "learning_rate": 2.9736517167390054e-05, "loss": 0.0309, "step": 104190 }, { "epoch": 2.0822509092362416, "grad_norm": 0.11513164639472961, "learning_rate": 2.972707011611696e-05, "loss": 0.0201, "step": 104200 }, { "epoch": 2.0824507413772433, "grad_norm": 0.10140923410654068, "learning_rate": 2.9717623930866178e-05, "loss": 0.0289, "step": 104210 }, { "epoch": 2.0826505735182446, "grad_norm": 0.13447174429893494, "learning_rate": 2.970817861204126e-05, "loss": 0.0258, "step": 104220 }, { "epoch": 2.0828504056592463, "grad_norm": 0.10139723867177963, "learning_rate": 2.969873416004564e-05, "loss": 0.0335, "step": 104230 }, { "epoch": 2.083050237800248, "grad_norm": 0.14832362532615662, "learning_rate": 2.968929057528282e-05, "loss": 0.0363, "step": 104240 }, { "epoch": 2.0832500699412493, "grad_norm": 0.13433469831943512, "learning_rate": 2.9679847858156206e-05, "loss": 0.0227, "step": 104250 }, { "epoch": 2.083449902082251, "grad_norm": 0.14789173007011414, "learning_rate": 2.9670406009069147e-05, "loss": 0.0285, "step": 104260 }, { "epoch": 2.0836497342232523, "grad_norm": 0.10252181440591812, "learning_rate": 2.9660965028424986e-05, "loss": 0.0222, "step": 104270 }, { "epoch": 2.083849566364254, "grad_norm": 0.11633872985839844, "learning_rate": 2.965152491662706e-05, "loss": 0.0287, "step": 104280 }, { "epoch": 2.0840493985052557, "grad_norm": 0.1403530389070511, "learning_rate": 2.9642085674078595e-05, "loss": 0.0245, "step": 104290 }, { "epoch": 2.084249230646257, "grad_norm": 0.11796127259731293, "learning_rate": 2.9632647301182842e-05, "loss": 0.025, "step": 104300 }, { "epoch": 2.0844490627872587, "grad_norm": 0.1585037261247635, "learning_rate": 2.9623209798342972e-05, "loss": 0.0329, "step": 104310 }, { "epoch": 2.0846488949282604, "grad_norm": 0.10830790549516678, "learning_rate": 2.9613773165962134e-05, "loss": 0.0229, "step": 104320 }, { "epoch": 2.0848487270692617, "grad_norm": 0.11147245019674301, "learning_rate": 2.960433740444348e-05, "loss": 0.0239, "step": 104330 }, { "epoch": 2.0850485592102634, "grad_norm": 0.10404915362596512, "learning_rate": 2.959490251419005e-05, "loss": 0.0204, "step": 104340 }, { "epoch": 2.085248391351265, "grad_norm": 0.1462077796459198, "learning_rate": 2.9585468495604906e-05, "loss": 0.0263, "step": 104350 }, { "epoch": 2.0854482234922664, "grad_norm": 0.13009880483150482, "learning_rate": 2.9576035349091024e-05, "loss": 0.0266, "step": 104360 }, { "epoch": 2.085648055633268, "grad_norm": 0.1395096778869629, "learning_rate": 2.9566603075051413e-05, "loss": 0.0259, "step": 104370 }, { "epoch": 2.0858478877742694, "grad_norm": 0.10518413037061691, "learning_rate": 2.9557171673888984e-05, "loss": 0.0216, "step": 104380 }, { "epoch": 2.086047719915271, "grad_norm": 0.11615028232336044, "learning_rate": 2.9547741146006614e-05, "loss": 0.0253, "step": 104390 }, { "epoch": 2.086247552056273, "grad_norm": 0.15691311657428741, "learning_rate": 2.9538311491807185e-05, "loss": 0.0245, "step": 104400 }, { "epoch": 2.086447384197274, "grad_norm": 0.11890330165624619, "learning_rate": 2.952888271169348e-05, "loss": 0.0349, "step": 104410 }, { "epoch": 2.086647216338276, "grad_norm": 0.12407425045967102, "learning_rate": 2.9519454806068305e-05, "loss": 0.0282, "step": 104420 }, { "epoch": 2.0868470484792776, "grad_norm": 0.16534999012947083, "learning_rate": 2.9510027775334408e-05, "loss": 0.0339, "step": 104430 }, { "epoch": 2.087046880620279, "grad_norm": 0.14381341636180878, "learning_rate": 2.9500601619894475e-05, "loss": 0.0237, "step": 104440 }, { "epoch": 2.0872467127612806, "grad_norm": 0.11566125601530075, "learning_rate": 2.9491176340151173e-05, "loss": 0.0267, "step": 104450 }, { "epoch": 2.087446544902282, "grad_norm": 0.1412765234708786, "learning_rate": 2.948175193650715e-05, "loss": 0.0295, "step": 104460 }, { "epoch": 2.0876463770432836, "grad_norm": 0.125957190990448, "learning_rate": 2.9472328409365e-05, "loss": 0.0328, "step": 104470 }, { "epoch": 2.0878462091842853, "grad_norm": 0.10673067718744278, "learning_rate": 2.9462905759127257e-05, "loss": 0.0325, "step": 104480 }, { "epoch": 2.0880460413252866, "grad_norm": 0.11411052942276001, "learning_rate": 2.945348398619647e-05, "loss": 0.0289, "step": 104490 }, { "epoch": 2.0882458734662883, "grad_norm": 0.13371804356575012, "learning_rate": 2.9444063090975083e-05, "loss": 0.0243, "step": 104500 }, { "epoch": 2.08844570560729, "grad_norm": 0.11621425300836563, "learning_rate": 2.9434643073865558e-05, "loss": 0.0241, "step": 104510 }, { "epoch": 2.0886455377482913, "grad_norm": 0.09748510271310806, "learning_rate": 2.9425223935270318e-05, "loss": 0.027, "step": 104520 }, { "epoch": 2.088845369889293, "grad_norm": 0.146749809384346, "learning_rate": 2.94158056755917e-05, "loss": 0.0258, "step": 104530 }, { "epoch": 2.0890452020302948, "grad_norm": 0.10886640846729279, "learning_rate": 2.9406388295232047e-05, "loss": 0.0249, "step": 104540 }, { "epoch": 2.089245034171296, "grad_norm": 0.1198173388838768, "learning_rate": 2.9396971794593663e-05, "loss": 0.023, "step": 104550 }, { "epoch": 2.0894448663122978, "grad_norm": 0.0946907326579094, "learning_rate": 2.9387556174078795e-05, "loss": 0.026, "step": 104560 }, { "epoch": 2.089644698453299, "grad_norm": 0.1275613009929657, "learning_rate": 2.937814143408967e-05, "loss": 0.0186, "step": 104570 }, { "epoch": 2.0898445305943008, "grad_norm": 0.08870251476764679, "learning_rate": 2.9368727575028445e-05, "loss": 0.0243, "step": 104580 }, { "epoch": 2.0900443627353025, "grad_norm": 0.11893415451049805, "learning_rate": 2.9359314597297267e-05, "loss": 0.0297, "step": 104590 }, { "epoch": 2.0902441948763038, "grad_norm": 0.1089542955160141, "learning_rate": 2.9349902501298276e-05, "loss": 0.029, "step": 104600 }, { "epoch": 2.0904440270173055, "grad_norm": 0.11184120923280716, "learning_rate": 2.9340491287433502e-05, "loss": 0.0276, "step": 104610 }, { "epoch": 2.090643859158307, "grad_norm": 0.10498147457838058, "learning_rate": 2.9331080956104993e-05, "loss": 0.027, "step": 104620 }, { "epoch": 2.0908436912993085, "grad_norm": 0.09673222154378891, "learning_rate": 2.932167150771472e-05, "loss": 0.0279, "step": 104630 }, { "epoch": 2.09104352344031, "grad_norm": 0.10275035351514816, "learning_rate": 2.9312262942664664e-05, "loss": 0.0192, "step": 104640 }, { "epoch": 2.0912433555813115, "grad_norm": 0.10234872251749039, "learning_rate": 2.9302855261356743e-05, "loss": 0.0232, "step": 104650 }, { "epoch": 2.091443187722313, "grad_norm": 0.11714058369398117, "learning_rate": 2.9293448464192806e-05, "loss": 0.0368, "step": 104660 }, { "epoch": 2.091643019863315, "grad_norm": 0.09661959856748581, "learning_rate": 2.9284042551574724e-05, "loss": 0.024, "step": 104670 }, { "epoch": 2.091842852004316, "grad_norm": 0.10133440047502518, "learning_rate": 2.9274637523904268e-05, "loss": 0.0235, "step": 104680 }, { "epoch": 2.092042684145318, "grad_norm": 0.12525109946727753, "learning_rate": 2.9265233381583236e-05, "loss": 0.0244, "step": 104690 }, { "epoch": 2.0922425162863196, "grad_norm": 0.11720171570777893, "learning_rate": 2.925583012501335e-05, "loss": 0.0242, "step": 104700 }, { "epoch": 2.092442348427321, "grad_norm": 0.09582718461751938, "learning_rate": 2.9246427754596285e-05, "loss": 0.0239, "step": 104710 }, { "epoch": 2.0926421805683226, "grad_norm": 0.14953601360321045, "learning_rate": 2.923702627073369e-05, "loss": 0.0215, "step": 104720 }, { "epoch": 2.0928420127093244, "grad_norm": 0.11195207387208939, "learning_rate": 2.9227625673827208e-05, "loss": 0.0275, "step": 104730 }, { "epoch": 2.0930418448503256, "grad_norm": 0.13841769099235535, "learning_rate": 2.921822596427839e-05, "loss": 0.0189, "step": 104740 }, { "epoch": 2.0932416769913273, "grad_norm": 0.10199065506458282, "learning_rate": 2.920882714248879e-05, "loss": 0.0274, "step": 104750 }, { "epoch": 2.0934415091323286, "grad_norm": 0.13572944700717926, "learning_rate": 2.9199429208859886e-05, "loss": 0.026, "step": 104760 }, { "epoch": 2.0936413412733303, "grad_norm": 0.1348118633031845, "learning_rate": 2.9190032163793157e-05, "loss": 0.0267, "step": 104770 }, { "epoch": 2.093841173414332, "grad_norm": 0.1490764170885086, "learning_rate": 2.918063600769003e-05, "loss": 0.0282, "step": 104780 }, { "epoch": 2.0940410055553333, "grad_norm": 0.14538682997226715, "learning_rate": 2.917124074095188e-05, "loss": 0.025, "step": 104790 }, { "epoch": 2.094240837696335, "grad_norm": 0.1446075737476349, "learning_rate": 2.9161846363980068e-05, "loss": 0.0273, "step": 104800 }, { "epoch": 2.094440669837337, "grad_norm": 0.11705052852630615, "learning_rate": 2.9152452877175897e-05, "loss": 0.0236, "step": 104810 }, { "epoch": 2.094640501978338, "grad_norm": 0.11914381384849548, "learning_rate": 2.9143060280940638e-05, "loss": 0.0269, "step": 104820 }, { "epoch": 2.09484033411934, "grad_norm": 0.14803697168827057, "learning_rate": 2.9133668575675525e-05, "loss": 0.0222, "step": 104830 }, { "epoch": 2.0950401662603415, "grad_norm": 0.09647019952535629, "learning_rate": 2.9124277761781784e-05, "loss": 0.0275, "step": 104840 }, { "epoch": 2.095239998401343, "grad_norm": 0.10317609459161758, "learning_rate": 2.911488783966052e-05, "loss": 0.0275, "step": 104850 }, { "epoch": 2.0954398305423445, "grad_norm": 0.11769171059131622, "learning_rate": 2.910549880971287e-05, "loss": 0.022, "step": 104860 }, { "epoch": 2.095639662683346, "grad_norm": 0.1289708912372589, "learning_rate": 2.909611067233994e-05, "loss": 0.0286, "step": 104870 }, { "epoch": 2.0958394948243475, "grad_norm": 0.12057317793369293, "learning_rate": 2.9086723427942787e-05, "loss": 0.0337, "step": 104880 }, { "epoch": 2.0960393269653492, "grad_norm": 0.11647684127092361, "learning_rate": 2.9077337076922363e-05, "loss": 0.0264, "step": 104890 }, { "epoch": 2.0962391591063505, "grad_norm": 0.14036564528942108, "learning_rate": 2.9067951619679644e-05, "loss": 0.0285, "step": 104900 }, { "epoch": 2.0964389912473522, "grad_norm": 0.11467462033033371, "learning_rate": 2.9058567056615614e-05, "loss": 0.0189, "step": 104910 }, { "epoch": 2.096638823388354, "grad_norm": 0.11054351180791855, "learning_rate": 2.904918338813112e-05, "loss": 0.0279, "step": 104920 }, { "epoch": 2.096838655529355, "grad_norm": 0.1287538707256317, "learning_rate": 2.9039800614627018e-05, "loss": 0.0201, "step": 104930 }, { "epoch": 2.097038487670357, "grad_norm": 0.11081007868051529, "learning_rate": 2.9030418736504128e-05, "loss": 0.0341, "step": 104940 }, { "epoch": 2.0972383198113587, "grad_norm": 0.11515391618013382, "learning_rate": 2.902103775416323e-05, "loss": 0.0325, "step": 104950 }, { "epoch": 2.09743815195236, "grad_norm": 0.14048068225383759, "learning_rate": 2.9011657668005054e-05, "loss": 0.0242, "step": 104960 }, { "epoch": 2.0976379840933617, "grad_norm": 0.09831603616476059, "learning_rate": 2.9002278478430312e-05, "loss": 0.0245, "step": 104970 }, { "epoch": 2.097837816234363, "grad_norm": 0.09245041012763977, "learning_rate": 2.8992900185839677e-05, "loss": 0.0213, "step": 104980 }, { "epoch": 2.0980376483753647, "grad_norm": 0.12847286462783813, "learning_rate": 2.898352279063371e-05, "loss": 0.027, "step": 104990 }, { "epoch": 2.0982374805163664, "grad_norm": 0.12363509833812714, "learning_rate": 2.8974146293213065e-05, "loss": 0.0248, "step": 105000 }, { "epoch": 2.0984373126573677, "grad_norm": 0.09784040600061417, "learning_rate": 2.8964770693978287e-05, "loss": 0.0335, "step": 105010 }, { "epoch": 2.0986371447983694, "grad_norm": 0.14388291537761688, "learning_rate": 2.8955395993329835e-05, "loss": 0.031, "step": 105020 }, { "epoch": 2.098836976939371, "grad_norm": 0.14278262853622437, "learning_rate": 2.8946022191668215e-05, "loss": 0.0319, "step": 105030 }, { "epoch": 2.0990368090803724, "grad_norm": 0.10201366245746613, "learning_rate": 2.8936649289393845e-05, "loss": 0.0293, "step": 105040 }, { "epoch": 2.099236641221374, "grad_norm": 0.12839056551456451, "learning_rate": 2.892727728690713e-05, "loss": 0.022, "step": 105050 }, { "epoch": 2.099436473362376, "grad_norm": 0.17214038968086243, "learning_rate": 2.89179061846084e-05, "loss": 0.0256, "step": 105060 }, { "epoch": 2.099636305503377, "grad_norm": 0.13642491400241852, "learning_rate": 2.8908535982898006e-05, "loss": 0.0267, "step": 105070 }, { "epoch": 2.099836137644379, "grad_norm": 0.1958087980747223, "learning_rate": 2.8899166682176205e-05, "loss": 0.0326, "step": 105080 }, { "epoch": 2.10003596978538, "grad_norm": 0.10477346926927567, "learning_rate": 2.8889798282843227e-05, "loss": 0.0298, "step": 105090 }, { "epoch": 2.100235801926382, "grad_norm": 0.1045021191239357, "learning_rate": 2.88804307852993e-05, "loss": 0.0261, "step": 105100 }, { "epoch": 2.1004356340673835, "grad_norm": 0.15389271080493927, "learning_rate": 2.8871064189944584e-05, "loss": 0.0226, "step": 105110 }, { "epoch": 2.100635466208385, "grad_norm": 0.09541446715593338, "learning_rate": 2.8861698497179178e-05, "loss": 0.0285, "step": 105120 }, { "epoch": 2.1008352983493865, "grad_norm": 0.13637623190879822, "learning_rate": 2.8852333707403156e-05, "loss": 0.0268, "step": 105130 }, { "epoch": 2.1010351304903883, "grad_norm": 0.09454963356256485, "learning_rate": 2.884296982101663e-05, "loss": 0.0253, "step": 105140 }, { "epoch": 2.1012349626313895, "grad_norm": 0.1468104124069214, "learning_rate": 2.8833606838419548e-05, "loss": 0.0237, "step": 105150 }, { "epoch": 2.1014347947723913, "grad_norm": 0.13431395590305328, "learning_rate": 2.8824244760011887e-05, "loss": 0.0281, "step": 105160 }, { "epoch": 2.101634626913393, "grad_norm": 0.1078495979309082, "learning_rate": 2.8814883586193597e-05, "loss": 0.0256, "step": 105170 }, { "epoch": 2.1018344590543943, "grad_norm": 0.12459852546453476, "learning_rate": 2.8805523317364563e-05, "loss": 0.0171, "step": 105180 }, { "epoch": 2.102034291195396, "grad_norm": 0.12148961424827576, "learning_rate": 2.879616395392464e-05, "loss": 0.0223, "step": 105190 }, { "epoch": 2.1022341233363973, "grad_norm": 0.11261415481567383, "learning_rate": 2.8786805496273633e-05, "loss": 0.0244, "step": 105200 }, { "epoch": 2.102433955477399, "grad_norm": 0.12163206189870834, "learning_rate": 2.8777447944811353e-05, "loss": 0.0323, "step": 105210 }, { "epoch": 2.1026337876184007, "grad_norm": 0.0983857735991478, "learning_rate": 2.8768091299937462e-05, "loss": 0.03, "step": 105220 }, { "epoch": 2.102833619759402, "grad_norm": 0.12022463232278824, "learning_rate": 2.875873556205173e-05, "loss": 0.0318, "step": 105230 }, { "epoch": 2.1030334519004037, "grad_norm": 0.10897859185934067, "learning_rate": 2.874938073155381e-05, "loss": 0.0288, "step": 105240 }, { "epoch": 2.1032332840414054, "grad_norm": 0.1661709100008011, "learning_rate": 2.874002680884329e-05, "loss": 0.0313, "step": 105250 }, { "epoch": 2.1034331161824067, "grad_norm": 0.09691973030567169, "learning_rate": 2.873067379431975e-05, "loss": 0.0237, "step": 105260 }, { "epoch": 2.1036329483234084, "grad_norm": 0.11476314067840576, "learning_rate": 2.8721321688382775e-05, "loss": 0.0258, "step": 105270 }, { "epoch": 2.1038327804644097, "grad_norm": 0.17374449968338013, "learning_rate": 2.871197049143186e-05, "loss": 0.0247, "step": 105280 }, { "epoch": 2.1040326126054114, "grad_norm": 0.09218885004520416, "learning_rate": 2.8702620203866438e-05, "loss": 0.0288, "step": 105290 }, { "epoch": 2.104232444746413, "grad_norm": 0.11872723698616028, "learning_rate": 2.8693270826085937e-05, "loss": 0.0352, "step": 105300 }, { "epoch": 2.1044322768874144, "grad_norm": 0.14357087016105652, "learning_rate": 2.8683922358489814e-05, "loss": 0.0316, "step": 105310 }, { "epoch": 2.104632109028416, "grad_norm": 0.13332127034664154, "learning_rate": 2.8674574801477337e-05, "loss": 0.0195, "step": 105320 }, { "epoch": 2.104831941169418, "grad_norm": 0.1996408998966217, "learning_rate": 2.8665228155447854e-05, "loss": 0.0384, "step": 105330 }, { "epoch": 2.105031773310419, "grad_norm": 0.17371512949466705, "learning_rate": 2.8655882420800634e-05, "loss": 0.0297, "step": 105340 }, { "epoch": 2.105231605451421, "grad_norm": 0.11556904017925262, "learning_rate": 2.864653759793491e-05, "loss": 0.0287, "step": 105350 }, { "epoch": 2.1054314375924226, "grad_norm": 0.12324418872594833, "learning_rate": 2.8637193687249863e-05, "loss": 0.0306, "step": 105360 }, { "epoch": 2.105631269733424, "grad_norm": 0.11534661054611206, "learning_rate": 2.862785068914465e-05, "loss": 0.0293, "step": 105370 }, { "epoch": 2.1058311018744256, "grad_norm": 0.16080190241336823, "learning_rate": 2.8618508604018428e-05, "loss": 0.027, "step": 105380 }, { "epoch": 2.106030934015427, "grad_norm": 0.1674206554889679, "learning_rate": 2.8609167432270197e-05, "loss": 0.0255, "step": 105390 }, { "epoch": 2.1062307661564286, "grad_norm": 0.1125388815999031, "learning_rate": 2.8599827174299056e-05, "loss": 0.0286, "step": 105400 }, { "epoch": 2.1064305982974303, "grad_norm": 0.0922832116484642, "learning_rate": 2.8590487830504e-05, "loss": 0.0187, "step": 105410 }, { "epoch": 2.1066304304384316, "grad_norm": 0.12197384238243103, "learning_rate": 2.8581149401283957e-05, "loss": 0.027, "step": 105420 }, { "epoch": 2.1068302625794333, "grad_norm": 0.11134099960327148, "learning_rate": 2.8571811887037863e-05, "loss": 0.02, "step": 105430 }, { "epoch": 2.107030094720435, "grad_norm": 0.10173629969358444, "learning_rate": 2.8562475288164605e-05, "loss": 0.0248, "step": 105440 }, { "epoch": 2.1072299268614363, "grad_norm": 0.11658351868391037, "learning_rate": 2.8553139605063017e-05, "loss": 0.0305, "step": 105450 }, { "epoch": 2.107429759002438, "grad_norm": 0.1076015904545784, "learning_rate": 2.8543804838131905e-05, "loss": 0.0226, "step": 105460 }, { "epoch": 2.1076295911434393, "grad_norm": 0.11173134297132492, "learning_rate": 2.8534470987770033e-05, "loss": 0.0278, "step": 105470 }, { "epoch": 2.107829423284441, "grad_norm": 0.15799863636493683, "learning_rate": 2.8525138054376127e-05, "loss": 0.0303, "step": 105480 }, { "epoch": 2.1080292554254427, "grad_norm": 0.12523122131824493, "learning_rate": 2.8515806038348873e-05, "loss": 0.0253, "step": 105490 }, { "epoch": 2.108229087566444, "grad_norm": 0.18366508185863495, "learning_rate": 2.8506474940086924e-05, "loss": 0.0216, "step": 105500 }, { "epoch": 2.1084289197074457, "grad_norm": 0.14165064692497253, "learning_rate": 2.8497144759988893e-05, "loss": 0.0295, "step": 105510 }, { "epoch": 2.1086287518484474, "grad_norm": 0.11991790682077408, "learning_rate": 2.848781549845332e-05, "loss": 0.0373, "step": 105520 }, { "epoch": 2.1088285839894487, "grad_norm": 0.1186724528670311, "learning_rate": 2.8478487155878724e-05, "loss": 0.0378, "step": 105530 }, { "epoch": 2.1090284161304504, "grad_norm": 0.15131719410419464, "learning_rate": 2.8469159732663665e-05, "loss": 0.0222, "step": 105540 }, { "epoch": 2.109228248271452, "grad_norm": 0.11470441520214081, "learning_rate": 2.8459833229206528e-05, "loss": 0.0293, "step": 105550 }, { "epoch": 2.1094280804124534, "grad_norm": 0.15686771273612976, "learning_rate": 2.8450507645905744e-05, "loss": 0.0328, "step": 105560 }, { "epoch": 2.109627912553455, "grad_norm": 0.0949614867568016, "learning_rate": 2.8441182983159686e-05, "loss": 0.023, "step": 105570 }, { "epoch": 2.1098277446944564, "grad_norm": 0.1325305700302124, "learning_rate": 2.8431859241366686e-05, "loss": 0.0248, "step": 105580 }, { "epoch": 2.110027576835458, "grad_norm": 0.13123725354671478, "learning_rate": 2.842253642092504e-05, "loss": 0.0195, "step": 105590 }, { "epoch": 2.11022740897646, "grad_norm": 0.13096436858177185, "learning_rate": 2.8413214522233e-05, "loss": 0.0516, "step": 105600 }, { "epoch": 2.110427241117461, "grad_norm": 0.1642799824476242, "learning_rate": 2.8403893545688798e-05, "loss": 0.0236, "step": 105610 }, { "epoch": 2.110627073258463, "grad_norm": 0.08955695480108261, "learning_rate": 2.8394573491690548e-05, "loss": 0.0227, "step": 105620 }, { "epoch": 2.1108269053994646, "grad_norm": 0.09363190829753876, "learning_rate": 2.838525436063645e-05, "loss": 0.0273, "step": 105630 }, { "epoch": 2.111026737540466, "grad_norm": 0.13892799615859985, "learning_rate": 2.8375936152924598e-05, "loss": 0.0225, "step": 105640 }, { "epoch": 2.1112265696814676, "grad_norm": 0.12980620563030243, "learning_rate": 2.8366618868953004e-05, "loss": 0.0246, "step": 105650 }, { "epoch": 2.1114264018224693, "grad_norm": 0.13302288949489594, "learning_rate": 2.83573025091197e-05, "loss": 0.0234, "step": 105660 }, { "epoch": 2.1116262339634706, "grad_norm": 0.10868746042251587, "learning_rate": 2.8347987073822684e-05, "loss": 0.0221, "step": 105670 }, { "epoch": 2.1118260661044723, "grad_norm": 0.12693805992603302, "learning_rate": 2.833867256345991e-05, "loss": 0.02, "step": 105680 }, { "epoch": 2.1120258982454736, "grad_norm": 0.12854842841625214, "learning_rate": 2.8329358978429227e-05, "loss": 0.0275, "step": 105690 }, { "epoch": 2.1122257303864753, "grad_norm": 0.12209739536046982, "learning_rate": 2.8320046319128525e-05, "loss": 0.0269, "step": 105700 }, { "epoch": 2.112425562527477, "grad_norm": 0.11044463515281677, "learning_rate": 2.831073458595561e-05, "loss": 0.0311, "step": 105710 }, { "epoch": 2.1126253946684783, "grad_norm": 0.14710769057273865, "learning_rate": 2.8301423779308268e-05, "loss": 0.0248, "step": 105720 }, { "epoch": 2.11282522680948, "grad_norm": 0.1174406111240387, "learning_rate": 2.829211389958425e-05, "loss": 0.0282, "step": 105730 }, { "epoch": 2.1130250589504818, "grad_norm": 0.1757165640592575, "learning_rate": 2.828280494718124e-05, "loss": 0.0251, "step": 105740 }, { "epoch": 2.113224891091483, "grad_norm": 0.10275224596261978, "learning_rate": 2.8273496922496907e-05, "loss": 0.0214, "step": 105750 }, { "epoch": 2.1134247232324848, "grad_norm": 0.11610832065343857, "learning_rate": 2.826418982592887e-05, "loss": 0.031, "step": 105760 }, { "epoch": 2.1136245553734865, "grad_norm": 0.11127708852291107, "learning_rate": 2.825488365787472e-05, "loss": 0.0247, "step": 105770 }, { "epoch": 2.1138243875144878, "grad_norm": 0.1253117173910141, "learning_rate": 2.824557841873201e-05, "loss": 0.0238, "step": 105780 }, { "epoch": 2.1140242196554895, "grad_norm": 0.11735806614160538, "learning_rate": 2.823627410889821e-05, "loss": 0.0219, "step": 105790 }, { "epoch": 2.1142240517964908, "grad_norm": 0.11455177515745163, "learning_rate": 2.8226970728770774e-05, "loss": 0.0258, "step": 105800 }, { "epoch": 2.1144238839374925, "grad_norm": 0.10624600946903229, "learning_rate": 2.8217668278747188e-05, "loss": 0.033, "step": 105810 }, { "epoch": 2.114623716078494, "grad_norm": 0.13749460875988007, "learning_rate": 2.820836675922478e-05, "loss": 0.0248, "step": 105820 }, { "epoch": 2.1148235482194955, "grad_norm": 0.16699077188968658, "learning_rate": 2.819906617060092e-05, "loss": 0.0193, "step": 105830 }, { "epoch": 2.115023380360497, "grad_norm": 0.11725309491157532, "learning_rate": 2.81897665132729e-05, "loss": 0.0206, "step": 105840 }, { "epoch": 2.115223212501499, "grad_norm": 0.1252865046262741, "learning_rate": 2.8180467787637976e-05, "loss": 0.0223, "step": 105850 }, { "epoch": 2.1154230446425, "grad_norm": 0.1432650238275528, "learning_rate": 2.8171169994093392e-05, "loss": 0.0292, "step": 105860 }, { "epoch": 2.115622876783502, "grad_norm": 0.1302802711725235, "learning_rate": 2.816187313303633e-05, "loss": 0.0339, "step": 105870 }, { "epoch": 2.1158227089245036, "grad_norm": 0.20142588019371033, "learning_rate": 2.815257720486394e-05, "loss": 0.031, "step": 105880 }, { "epoch": 2.116022541065505, "grad_norm": 0.11933883279561996, "learning_rate": 2.8143282209973287e-05, "loss": 0.0286, "step": 105890 }, { "epoch": 2.1162223732065066, "grad_norm": 0.1049911305308342, "learning_rate": 2.8133988148761482e-05, "loss": 0.0251, "step": 105900 }, { "epoch": 2.116422205347508, "grad_norm": 0.12408510595560074, "learning_rate": 2.8124695021625546e-05, "loss": 0.0323, "step": 105910 }, { "epoch": 2.1166220374885096, "grad_norm": 0.12400021404027939, "learning_rate": 2.8115402828962445e-05, "loss": 0.0252, "step": 105920 }, { "epoch": 2.1168218696295114, "grad_norm": 0.11658938229084015, "learning_rate": 2.8106111571169104e-05, "loss": 0.0318, "step": 105930 }, { "epoch": 2.1170217017705126, "grad_norm": 0.11595000326633453, "learning_rate": 2.8096821248642503e-05, "loss": 0.0246, "step": 105940 }, { "epoch": 2.1172215339115144, "grad_norm": 0.13886423408985138, "learning_rate": 2.808753186177944e-05, "loss": 0.0222, "step": 105950 }, { "epoch": 2.117421366052516, "grad_norm": 0.13216815888881683, "learning_rate": 2.8078243410976757e-05, "loss": 0.0281, "step": 105960 }, { "epoch": 2.1176211981935174, "grad_norm": 0.1456194519996643, "learning_rate": 2.8068955896631245e-05, "loss": 0.0304, "step": 105970 }, { "epoch": 2.117821030334519, "grad_norm": 0.13949576020240784, "learning_rate": 2.8059669319139646e-05, "loss": 0.0235, "step": 105980 }, { "epoch": 2.1180208624755203, "grad_norm": 0.15492019057273865, "learning_rate": 2.8050383678898674e-05, "loss": 0.0369, "step": 105990 }, { "epoch": 2.118220694616522, "grad_norm": 0.09792882949113846, "learning_rate": 2.804109897630498e-05, "loss": 0.0223, "step": 106000 }, { "epoch": 2.118420526757524, "grad_norm": 0.1878613829612732, "learning_rate": 2.8031815211755224e-05, "loss": 0.0337, "step": 106010 }, { "epoch": 2.118620358898525, "grad_norm": 0.09806591272354126, "learning_rate": 2.8022532385645928e-05, "loss": 0.0275, "step": 106020 }, { "epoch": 2.118820191039527, "grad_norm": 0.1344430297613144, "learning_rate": 2.801325049837369e-05, "loss": 0.0324, "step": 106030 }, { "epoch": 2.1190200231805285, "grad_norm": 0.10905048251152039, "learning_rate": 2.800396955033502e-05, "loss": 0.0209, "step": 106040 }, { "epoch": 2.11921985532153, "grad_norm": 0.1527395248413086, "learning_rate": 2.7994689541926345e-05, "loss": 0.0252, "step": 106050 }, { "epoch": 2.1194196874625315, "grad_norm": 0.15658791363239288, "learning_rate": 2.7985410473544105e-05, "loss": 0.0273, "step": 106060 }, { "epoch": 2.1196195196035332, "grad_norm": 0.11071095615625381, "learning_rate": 2.7976132345584664e-05, "loss": 0.0231, "step": 106070 }, { "epoch": 2.1198193517445345, "grad_norm": 0.09248562157154083, "learning_rate": 2.796685515844444e-05, "loss": 0.0202, "step": 106080 }, { "epoch": 2.1200191838855362, "grad_norm": 0.12813112139701843, "learning_rate": 2.7957578912519654e-05, "loss": 0.0375, "step": 106090 }, { "epoch": 2.1202190160265375, "grad_norm": 0.09919331967830658, "learning_rate": 2.7948303608206617e-05, "loss": 0.0245, "step": 106100 }, { "epoch": 2.1204188481675392, "grad_norm": 0.15083003044128418, "learning_rate": 2.793902924590154e-05, "loss": 0.0253, "step": 106110 }, { "epoch": 2.120618680308541, "grad_norm": 0.14505861699581146, "learning_rate": 2.7929755826000592e-05, "loss": 0.024, "step": 106120 }, { "epoch": 2.1208185124495422, "grad_norm": 0.14887823164463043, "learning_rate": 2.792048334889994e-05, "loss": 0.0182, "step": 106130 }, { "epoch": 2.121018344590544, "grad_norm": 0.14554955065250397, "learning_rate": 2.791121181499568e-05, "loss": 0.0209, "step": 106140 }, { "epoch": 2.1212181767315457, "grad_norm": 0.12412051111459732, "learning_rate": 2.7901941224683892e-05, "loss": 0.0341, "step": 106150 }, { "epoch": 2.121418008872547, "grad_norm": 0.13842065632343292, "learning_rate": 2.7892671578360536e-05, "loss": 0.026, "step": 106160 }, { "epoch": 2.1216178410135487, "grad_norm": 0.12944960594177246, "learning_rate": 2.788340287642166e-05, "loss": 0.0206, "step": 106170 }, { "epoch": 2.1218176731545504, "grad_norm": 0.1017085388302803, "learning_rate": 2.78741351192632e-05, "loss": 0.0257, "step": 106180 }, { "epoch": 2.1220175052955517, "grad_norm": 0.11895345151424408, "learning_rate": 2.786486830728103e-05, "loss": 0.0314, "step": 106190 }, { "epoch": 2.1222173374365534, "grad_norm": 0.12690262496471405, "learning_rate": 2.7855602440870993e-05, "loss": 0.0345, "step": 106200 }, { "epoch": 2.1224171695775547, "grad_norm": 0.11669867485761642, "learning_rate": 2.7846337520428982e-05, "loss": 0.0305, "step": 106210 }, { "epoch": 2.1226170017185564, "grad_norm": 0.11915143579244614, "learning_rate": 2.7837073546350716e-05, "loss": 0.022, "step": 106220 }, { "epoch": 2.122816833859558, "grad_norm": 0.11321142315864563, "learning_rate": 2.7827810519031956e-05, "loss": 0.0206, "step": 106230 }, { "epoch": 2.1230166660005594, "grad_norm": 0.13951903581619263, "learning_rate": 2.7818548438868396e-05, "loss": 0.0327, "step": 106240 }, { "epoch": 2.123216498141561, "grad_norm": 0.11694299429655075, "learning_rate": 2.7809287306255695e-05, "loss": 0.0252, "step": 106250 }, { "epoch": 2.123416330282563, "grad_norm": 0.1278896927833557, "learning_rate": 2.7800027121589478e-05, "loss": 0.0389, "step": 106260 }, { "epoch": 2.123616162423564, "grad_norm": 0.10906189680099487, "learning_rate": 2.7790767885265322e-05, "loss": 0.0382, "step": 106270 }, { "epoch": 2.123815994564566, "grad_norm": 0.1238468810915947, "learning_rate": 2.778150959767878e-05, "loss": 0.0283, "step": 106280 }, { "epoch": 2.124015826705567, "grad_norm": 0.08524338901042938, "learning_rate": 2.777225225922529e-05, "loss": 0.0277, "step": 106290 }, { "epoch": 2.124215658846569, "grad_norm": 0.1293666809797287, "learning_rate": 2.7762995870300368e-05, "loss": 0.0363, "step": 106300 }, { "epoch": 2.1244154909875705, "grad_norm": 0.16117316484451294, "learning_rate": 2.7753740431299437e-05, "loss": 0.024, "step": 106310 }, { "epoch": 2.124615323128572, "grad_norm": 0.12308913469314575, "learning_rate": 2.7744485942617828e-05, "loss": 0.0307, "step": 106320 }, { "epoch": 2.1248151552695735, "grad_norm": 0.12793266773223877, "learning_rate": 2.773523240465089e-05, "loss": 0.0349, "step": 106330 }, { "epoch": 2.1250149874105753, "grad_norm": 0.10556431859731674, "learning_rate": 2.7725979817793923e-05, "loss": 0.0174, "step": 106340 }, { "epoch": 2.1252148195515765, "grad_norm": 0.10897589474916458, "learning_rate": 2.7716728182442174e-05, "loss": 0.0257, "step": 106350 }, { "epoch": 2.1254146516925783, "grad_norm": 0.12902195751667023, "learning_rate": 2.770747749899087e-05, "loss": 0.0166, "step": 106360 }, { "epoch": 2.12561448383358, "grad_norm": 0.09822913259267807, "learning_rate": 2.7698227767835173e-05, "loss": 0.0339, "step": 106370 }, { "epoch": 2.1258143159745813, "grad_norm": 0.1161055713891983, "learning_rate": 2.768897898937022e-05, "loss": 0.0291, "step": 106380 }, { "epoch": 2.126014148115583, "grad_norm": 0.09909456223249435, "learning_rate": 2.7679731163991084e-05, "loss": 0.0279, "step": 106390 }, { "epoch": 2.1262139802565843, "grad_norm": 0.13283996284008026, "learning_rate": 2.7670484292092842e-05, "loss": 0.0343, "step": 106400 }, { "epoch": 2.126413812397586, "grad_norm": 0.14114083349704742, "learning_rate": 2.7661238374070508e-05, "loss": 0.0263, "step": 106410 }, { "epoch": 2.1266136445385877, "grad_norm": 0.1632159948348999, "learning_rate": 2.765199341031901e-05, "loss": 0.0245, "step": 106420 }, { "epoch": 2.126813476679589, "grad_norm": 0.10652263462543488, "learning_rate": 2.7642749401233283e-05, "loss": 0.0238, "step": 106430 }, { "epoch": 2.1270133088205907, "grad_norm": 0.14081226289272308, "learning_rate": 2.763350634720826e-05, "loss": 0.0261, "step": 106440 }, { "epoch": 2.1272131409615924, "grad_norm": 0.1314069628715515, "learning_rate": 2.7624264248638747e-05, "loss": 0.0191, "step": 106450 }, { "epoch": 2.1274129731025937, "grad_norm": 0.12557411193847656, "learning_rate": 2.7615023105919558e-05, "loss": 0.0224, "step": 106460 }, { "epoch": 2.1276128052435954, "grad_norm": 0.0940263643860817, "learning_rate": 2.7605782919445438e-05, "loss": 0.0194, "step": 106470 }, { "epoch": 2.127812637384597, "grad_norm": 0.1120312362909317, "learning_rate": 2.7596543689611176e-05, "loss": 0.035, "step": 106480 }, { "epoch": 2.1280124695255984, "grad_norm": 0.16353651881217957, "learning_rate": 2.758730541681138e-05, "loss": 0.0201, "step": 106490 }, { "epoch": 2.1282123016666, "grad_norm": 0.1293257176876068, "learning_rate": 2.7578068101440735e-05, "loss": 0.0263, "step": 106500 }, { "epoch": 2.1284121338076014, "grad_norm": 0.15748648345470428, "learning_rate": 2.756883174389383e-05, "loss": 0.0367, "step": 106510 }, { "epoch": 2.128611965948603, "grad_norm": 0.11795469373464584, "learning_rate": 2.755959634456522e-05, "loss": 0.0299, "step": 106520 }, { "epoch": 2.128811798089605, "grad_norm": 0.10686948895454407, "learning_rate": 2.755036190384943e-05, "loss": 0.0289, "step": 106530 }, { "epoch": 2.129011630230606, "grad_norm": 0.16248942911624908, "learning_rate": 2.754112842214094e-05, "loss": 0.0316, "step": 106540 }, { "epoch": 2.129211462371608, "grad_norm": 0.12523207068443298, "learning_rate": 2.7531895899834202e-05, "loss": 0.0359, "step": 106550 }, { "epoch": 2.1294112945126096, "grad_norm": 0.16871178150177002, "learning_rate": 2.7522664337323568e-05, "loss": 0.0177, "step": 106560 }, { "epoch": 2.129611126653611, "grad_norm": 0.11001750826835632, "learning_rate": 2.751343373500343e-05, "loss": 0.0222, "step": 106570 }, { "epoch": 2.1298109587946126, "grad_norm": 0.13475923240184784, "learning_rate": 2.7504204093268117e-05, "loss": 0.0302, "step": 106580 }, { "epoch": 2.1300107909356143, "grad_norm": 0.1260700672864914, "learning_rate": 2.7494975412511857e-05, "loss": 0.0254, "step": 106590 }, { "epoch": 2.1302106230766156, "grad_norm": 0.11830798536539078, "learning_rate": 2.748574769312889e-05, "loss": 0.0304, "step": 106600 }, { "epoch": 2.1304104552176173, "grad_norm": 0.15231853723526, "learning_rate": 2.747652093551345e-05, "loss": 0.0292, "step": 106610 }, { "epoch": 2.1306102873586186, "grad_norm": 0.139007568359375, "learning_rate": 2.7467295140059652e-05, "loss": 0.0284, "step": 106620 }, { "epoch": 2.1308101194996203, "grad_norm": 0.1395408809185028, "learning_rate": 2.7458070307161603e-05, "loss": 0.0256, "step": 106630 }, { "epoch": 2.131009951640622, "grad_norm": 0.12042845785617828, "learning_rate": 2.7448846437213382e-05, "loss": 0.0275, "step": 106640 }, { "epoch": 2.1312097837816233, "grad_norm": 0.11261723190546036, "learning_rate": 2.7439623530609014e-05, "loss": 0.0266, "step": 106650 }, { "epoch": 2.131409615922625, "grad_norm": 0.12942199409008026, "learning_rate": 2.743040158774248e-05, "loss": 0.0279, "step": 106660 }, { "epoch": 2.1316094480636267, "grad_norm": 0.09029041975736618, "learning_rate": 2.7421180609007734e-05, "loss": 0.0258, "step": 106670 }, { "epoch": 2.131809280204628, "grad_norm": 0.10355541855096817, "learning_rate": 2.7411960594798697e-05, "loss": 0.0202, "step": 106680 }, { "epoch": 2.1320091123456297, "grad_norm": 0.14372718334197998, "learning_rate": 2.7402741545509157e-05, "loss": 0.031, "step": 106690 }, { "epoch": 2.1322089444866315, "grad_norm": 0.12885682284832, "learning_rate": 2.7393523461533017e-05, "loss": 0.0304, "step": 106700 }, { "epoch": 2.1324087766276327, "grad_norm": 0.130009263753891, "learning_rate": 2.7384306343264037e-05, "loss": 0.0299, "step": 106710 }, { "epoch": 2.1326086087686345, "grad_norm": 0.09997402876615524, "learning_rate": 2.7375090191095933e-05, "loss": 0.0192, "step": 106720 }, { "epoch": 2.1328084409096357, "grad_norm": 0.1414329558610916, "learning_rate": 2.7365875005422403e-05, "loss": 0.0267, "step": 106730 }, { "epoch": 2.1330082730506374, "grad_norm": 0.1031765267252922, "learning_rate": 2.7356660786637112e-05, "loss": 0.0269, "step": 106740 }, { "epoch": 2.133208105191639, "grad_norm": 0.14646463096141815, "learning_rate": 2.7347447535133685e-05, "loss": 0.0297, "step": 106750 }, { "epoch": 2.1334079373326404, "grad_norm": 0.13812901079654694, "learning_rate": 2.7338235251305678e-05, "loss": 0.0242, "step": 106760 }, { "epoch": 2.133607769473642, "grad_norm": 0.13522005081176758, "learning_rate": 2.7329023935546628e-05, "loss": 0.032, "step": 106770 }, { "epoch": 2.133807601614644, "grad_norm": 0.10792309790849686, "learning_rate": 2.731981358825003e-05, "loss": 0.0227, "step": 106780 }, { "epoch": 2.134007433755645, "grad_norm": 0.09399249404668808, "learning_rate": 2.7310604209809333e-05, "loss": 0.0262, "step": 106790 }, { "epoch": 2.134207265896647, "grad_norm": 0.1207382082939148, "learning_rate": 2.730139580061794e-05, "loss": 0.0285, "step": 106800 }, { "epoch": 2.1344070980376486, "grad_norm": 0.1131996437907219, "learning_rate": 2.7292188361069244e-05, "loss": 0.0234, "step": 106810 }, { "epoch": 2.13460693017865, "grad_norm": 0.1018066480755806, "learning_rate": 2.728298189155652e-05, "loss": 0.0275, "step": 106820 }, { "epoch": 2.1348067623196516, "grad_norm": 0.13494440913200378, "learning_rate": 2.727377639247306e-05, "loss": 0.0271, "step": 106830 }, { "epoch": 2.135006594460653, "grad_norm": 0.10979658365249634, "learning_rate": 2.726457186421214e-05, "loss": 0.0286, "step": 106840 }, { "epoch": 2.1352064266016546, "grad_norm": 0.12165208160877228, "learning_rate": 2.7255368307166967e-05, "loss": 0.0293, "step": 106850 }, { "epoch": 2.1354062587426563, "grad_norm": 0.10670065134763718, "learning_rate": 2.7246165721730653e-05, "loss": 0.0245, "step": 106860 }, { "epoch": 2.1356060908836576, "grad_norm": 0.20311051607131958, "learning_rate": 2.7236964108296325e-05, "loss": 0.0231, "step": 106870 }, { "epoch": 2.1358059230246593, "grad_norm": 0.11948902904987335, "learning_rate": 2.7227763467257112e-05, "loss": 0.0247, "step": 106880 }, { "epoch": 2.136005755165661, "grad_norm": 0.13473834097385406, "learning_rate": 2.7218563799005992e-05, "loss": 0.0307, "step": 106890 }, { "epoch": 2.1362055873066623, "grad_norm": 0.10294094681739807, "learning_rate": 2.720936510393597e-05, "loss": 0.0235, "step": 106900 }, { "epoch": 2.136405419447664, "grad_norm": 0.1122596263885498, "learning_rate": 2.720016738244e-05, "loss": 0.0292, "step": 106910 }, { "epoch": 2.1366052515886653, "grad_norm": 0.10706762224435806, "learning_rate": 2.7190970634911e-05, "loss": 0.0399, "step": 106920 }, { "epoch": 2.136805083729667, "grad_norm": 0.13002730906009674, "learning_rate": 2.7181774861741837e-05, "loss": 0.0273, "step": 106930 }, { "epoch": 2.1370049158706688, "grad_norm": 0.09376812726259232, "learning_rate": 2.7172580063325325e-05, "loss": 0.0178, "step": 106940 }, { "epoch": 2.13720474801167, "grad_norm": 0.13949038088321686, "learning_rate": 2.7163386240054277e-05, "loss": 0.0308, "step": 106950 }, { "epoch": 2.1374045801526718, "grad_norm": 0.21818901598453522, "learning_rate": 2.7154193392321388e-05, "loss": 0.0268, "step": 106960 }, { "epoch": 2.1376044122936735, "grad_norm": 0.16097497940063477, "learning_rate": 2.7145001520519396e-05, "loss": 0.0292, "step": 106970 }, { "epoch": 2.1378042444346748, "grad_norm": 0.12326528877019882, "learning_rate": 2.713581062504098e-05, "loss": 0.027, "step": 106980 }, { "epoch": 2.1380040765756765, "grad_norm": 0.11729839444160461, "learning_rate": 2.712662070627871e-05, "loss": 0.0241, "step": 106990 }, { "epoch": 2.1382039087166778, "grad_norm": 0.11962507665157318, "learning_rate": 2.7117431764625183e-05, "loss": 0.0218, "step": 107000 }, { "epoch": 2.1384037408576795, "grad_norm": 0.08467961102724075, "learning_rate": 2.7108243800472932e-05, "loss": 0.0304, "step": 107010 }, { "epoch": 2.138603572998681, "grad_norm": 0.123204305768013, "learning_rate": 2.709905681421446e-05, "loss": 0.0304, "step": 107020 }, { "epoch": 2.1388034051396825, "grad_norm": 0.11496039479970932, "learning_rate": 2.7089870806242203e-05, "loss": 0.029, "step": 107030 }, { "epoch": 2.139003237280684, "grad_norm": 0.11431512981653214, "learning_rate": 2.708068577694858e-05, "loss": 0.0212, "step": 107040 }, { "epoch": 2.139203069421686, "grad_norm": 0.12964311242103577, "learning_rate": 2.7071501726725957e-05, "loss": 0.0274, "step": 107050 }, { "epoch": 2.139402901562687, "grad_norm": 0.10182853043079376, "learning_rate": 2.7062318655966657e-05, "loss": 0.0279, "step": 107060 }, { "epoch": 2.139602733703689, "grad_norm": 0.10983709990978241, "learning_rate": 2.7053136565062976e-05, "loss": 0.0282, "step": 107070 }, { "epoch": 2.1398025658446906, "grad_norm": 0.10342789441347122, "learning_rate": 2.7043955454407156e-05, "loss": 0.0252, "step": 107080 }, { "epoch": 2.140002397985692, "grad_norm": 0.1548352688550949, "learning_rate": 2.703477532439137e-05, "loss": 0.0308, "step": 107090 }, { "epoch": 2.1402022301266936, "grad_norm": 0.15194171667099, "learning_rate": 2.7025596175407774e-05, "loss": 0.0286, "step": 107100 }, { "epoch": 2.140402062267695, "grad_norm": 0.09544822573661804, "learning_rate": 2.7016418007848543e-05, "loss": 0.0226, "step": 107110 }, { "epoch": 2.1406018944086966, "grad_norm": 0.08907479047775269, "learning_rate": 2.700724082210569e-05, "loss": 0.0188, "step": 107120 }, { "epoch": 2.1408017265496984, "grad_norm": 0.12051211297512054, "learning_rate": 2.699806461857127e-05, "loss": 0.0321, "step": 107130 }, { "epoch": 2.1410015586906996, "grad_norm": 0.10324369370937347, "learning_rate": 2.6988889397637278e-05, "loss": 0.0414, "step": 107140 }, { "epoch": 2.1412013908317014, "grad_norm": 0.15147054195404053, "learning_rate": 2.697971515969566e-05, "loss": 0.0261, "step": 107150 }, { "epoch": 2.141401222972703, "grad_norm": 0.14906728267669678, "learning_rate": 2.6970541905138314e-05, "loss": 0.0262, "step": 107160 }, { "epoch": 2.1416010551137044, "grad_norm": 0.13616324961185455, "learning_rate": 2.6961369634357114e-05, "loss": 0.0353, "step": 107170 }, { "epoch": 2.141800887254706, "grad_norm": 0.19137731194496155, "learning_rate": 2.6952198347743905e-05, "loss": 0.0218, "step": 107180 }, { "epoch": 2.142000719395708, "grad_norm": 0.14595425128936768, "learning_rate": 2.69430280456904e-05, "loss": 0.0299, "step": 107190 }, { "epoch": 2.142200551536709, "grad_norm": 0.14638745784759521, "learning_rate": 2.6933858728588395e-05, "loss": 0.033, "step": 107200 }, { "epoch": 2.142400383677711, "grad_norm": 0.0879979208111763, "learning_rate": 2.69246903968296e-05, "loss": 0.0161, "step": 107210 }, { "epoch": 2.142600215818712, "grad_norm": 0.13687629997730255, "learning_rate": 2.6915523050805612e-05, "loss": 0.0314, "step": 107220 }, { "epoch": 2.142800047959714, "grad_norm": 0.13605478405952454, "learning_rate": 2.690635669090805e-05, "loss": 0.0272, "step": 107230 }, { "epoch": 2.1429998801007155, "grad_norm": 0.087186798453331, "learning_rate": 2.6897191317528534e-05, "loss": 0.0244, "step": 107240 }, { "epoch": 2.143199712241717, "grad_norm": 0.12565234303474426, "learning_rate": 2.688802693105858e-05, "loss": 0.0258, "step": 107250 }, { "epoch": 2.1433995443827185, "grad_norm": 0.12473775446414948, "learning_rate": 2.6878863531889636e-05, "loss": 0.0276, "step": 107260 }, { "epoch": 2.1435993765237202, "grad_norm": 0.11998011916875839, "learning_rate": 2.686970112041317e-05, "loss": 0.0252, "step": 107270 }, { "epoch": 2.1437992086647215, "grad_norm": 0.14066097140312195, "learning_rate": 2.686053969702058e-05, "loss": 0.0234, "step": 107280 }, { "epoch": 2.1439990408057232, "grad_norm": 0.13718679547309875, "learning_rate": 2.6851379262103227e-05, "loss": 0.0277, "step": 107290 }, { "epoch": 2.144198872946725, "grad_norm": 0.13829264044761658, "learning_rate": 2.6842219816052423e-05, "loss": 0.0318, "step": 107300 }, { "epoch": 2.1443987050877262, "grad_norm": 0.13878677785396576, "learning_rate": 2.683306135925945e-05, "loss": 0.0307, "step": 107310 }, { "epoch": 2.144598537228728, "grad_norm": 0.11056217551231384, "learning_rate": 2.682390389211553e-05, "loss": 0.0192, "step": 107320 }, { "epoch": 2.1447983693697292, "grad_norm": 0.12184188514947891, "learning_rate": 2.6814747415011864e-05, "loss": 0.0223, "step": 107330 }, { "epoch": 2.144998201510731, "grad_norm": 0.11612248420715332, "learning_rate": 2.6805591928339598e-05, "loss": 0.0233, "step": 107340 }, { "epoch": 2.1451980336517327, "grad_norm": 0.09012528508901596, "learning_rate": 2.679643743248985e-05, "loss": 0.0253, "step": 107350 }, { "epoch": 2.145397865792734, "grad_norm": 0.11768725514411926, "learning_rate": 2.678728392785365e-05, "loss": 0.023, "step": 107360 }, { "epoch": 2.1455976979337357, "grad_norm": 0.12339706718921661, "learning_rate": 2.677813141482202e-05, "loss": 0.0205, "step": 107370 }, { "epoch": 2.1457975300747374, "grad_norm": 0.1588384062051773, "learning_rate": 2.6768979893785995e-05, "loss": 0.0329, "step": 107380 }, { "epoch": 2.1459973622157387, "grad_norm": 0.14349406957626343, "learning_rate": 2.6759829365136456e-05, "loss": 0.0235, "step": 107390 }, { "epoch": 2.1461971943567404, "grad_norm": 0.09353922307491302, "learning_rate": 2.6750679829264307e-05, "loss": 0.0327, "step": 107400 }, { "epoch": 2.146397026497742, "grad_norm": 0.18759408593177795, "learning_rate": 2.6741531286560407e-05, "loss": 0.0301, "step": 107410 }, { "epoch": 2.1465968586387434, "grad_norm": 0.12541773915290833, "learning_rate": 2.6732383737415566e-05, "loss": 0.0241, "step": 107420 }, { "epoch": 2.146796690779745, "grad_norm": 0.18574467301368713, "learning_rate": 2.6723237182220544e-05, "loss": 0.0348, "step": 107430 }, { "epoch": 2.1469965229207464, "grad_norm": 0.12506374716758728, "learning_rate": 2.671409162136607e-05, "loss": 0.0243, "step": 107440 }, { "epoch": 2.147196355061748, "grad_norm": 0.09738735109567642, "learning_rate": 2.6704947055242846e-05, "loss": 0.0262, "step": 107450 }, { "epoch": 2.14739618720275, "grad_norm": 0.10371289402246475, "learning_rate": 2.669580348424145e-05, "loss": 0.0252, "step": 107460 }, { "epoch": 2.147596019343751, "grad_norm": 0.10955750197172165, "learning_rate": 2.6686660908752535e-05, "loss": 0.0251, "step": 107470 }, { "epoch": 2.147795851484753, "grad_norm": 0.10759297013282776, "learning_rate": 2.6677519329166655e-05, "loss": 0.0248, "step": 107480 }, { "epoch": 2.1479956836257545, "grad_norm": 0.12409649044275284, "learning_rate": 2.666837874587429e-05, "loss": 0.021, "step": 107490 }, { "epoch": 2.148195515766756, "grad_norm": 0.10505448281764984, "learning_rate": 2.66592391592659e-05, "loss": 0.0184, "step": 107500 }, { "epoch": 2.1483953479077575, "grad_norm": 0.1585080772638321, "learning_rate": 2.6650100569731983e-05, "loss": 0.0277, "step": 107510 }, { "epoch": 2.1485951800487593, "grad_norm": 0.13492341339588165, "learning_rate": 2.6640962977662847e-05, "loss": 0.0325, "step": 107520 }, { "epoch": 2.1487950121897605, "grad_norm": 0.11987221240997314, "learning_rate": 2.6631826383448867e-05, "loss": 0.0316, "step": 107530 }, { "epoch": 2.1489948443307623, "grad_norm": 0.13165903091430664, "learning_rate": 2.6622690787480325e-05, "loss": 0.0235, "step": 107540 }, { "epoch": 2.1491946764717635, "grad_norm": 0.11440730094909668, "learning_rate": 2.6613556190147494e-05, "loss": 0.0239, "step": 107550 }, { "epoch": 2.1493945086127653, "grad_norm": 0.10575845837593079, "learning_rate": 2.6604422591840578e-05, "loss": 0.0227, "step": 107560 }, { "epoch": 2.149594340753767, "grad_norm": 0.16565299034118652, "learning_rate": 2.659528999294975e-05, "loss": 0.0244, "step": 107570 }, { "epoch": 2.1497941728947683, "grad_norm": 0.1002001091837883, "learning_rate": 2.658615839386516e-05, "loss": 0.0177, "step": 107580 }, { "epoch": 2.14999400503577, "grad_norm": 0.16500329971313477, "learning_rate": 2.6577027794976827e-05, "loss": 0.0316, "step": 107590 }, { "epoch": 2.1501938371767717, "grad_norm": 0.0641099065542221, "learning_rate": 2.6567898196674855e-05, "loss": 0.0255, "step": 107600 }, { "epoch": 2.150393669317773, "grad_norm": 0.11840666085481644, "learning_rate": 2.6558769599349254e-05, "loss": 0.0278, "step": 107610 }, { "epoch": 2.1505935014587747, "grad_norm": 0.12302125990390778, "learning_rate": 2.654964200338993e-05, "loss": 0.0238, "step": 107620 }, { "epoch": 2.150793333599776, "grad_norm": 0.10259097069501877, "learning_rate": 2.6540515409186816e-05, "loss": 0.0223, "step": 107630 }, { "epoch": 2.1509931657407777, "grad_norm": 0.1091749370098114, "learning_rate": 2.6531389817129772e-05, "loss": 0.0213, "step": 107640 }, { "epoch": 2.1511929978817794, "grad_norm": 0.10633625090122223, "learning_rate": 2.652226522760868e-05, "loss": 0.0183, "step": 107650 }, { "epoch": 2.1513928300227807, "grad_norm": 0.21102717518806458, "learning_rate": 2.6513141641013268e-05, "loss": 0.0288, "step": 107660 }, { "epoch": 2.1515926621637824, "grad_norm": 0.1468137800693512, "learning_rate": 2.65040190577333e-05, "loss": 0.0222, "step": 107670 }, { "epoch": 2.151792494304784, "grad_norm": 0.12075494974851608, "learning_rate": 2.649489747815847e-05, "loss": 0.0275, "step": 107680 }, { "epoch": 2.1519923264457854, "grad_norm": 0.11401300132274628, "learning_rate": 2.648577690267844e-05, "loss": 0.0199, "step": 107690 }, { "epoch": 2.152192158586787, "grad_norm": 0.09086744487285614, "learning_rate": 2.647665733168283e-05, "loss": 0.0276, "step": 107700 }, { "epoch": 2.152391990727789, "grad_norm": 0.09367095679044724, "learning_rate": 2.6467538765561196e-05, "loss": 0.0325, "step": 107710 }, { "epoch": 2.15259182286879, "grad_norm": 0.11493184417486191, "learning_rate": 2.645842120470308e-05, "loss": 0.0239, "step": 107720 }, { "epoch": 2.152791655009792, "grad_norm": 0.11354567855596542, "learning_rate": 2.6449304649497964e-05, "loss": 0.0245, "step": 107730 }, { "epoch": 2.152991487150793, "grad_norm": 0.09891090542078018, "learning_rate": 2.644018910033529e-05, "loss": 0.0221, "step": 107740 }, { "epoch": 2.153191319291795, "grad_norm": 0.1616707742214203, "learning_rate": 2.6431074557604473e-05, "loss": 0.0219, "step": 107750 }, { "epoch": 2.1533911514327966, "grad_norm": 0.12789998948574066, "learning_rate": 2.6421961021694837e-05, "loss": 0.0239, "step": 107760 }, { "epoch": 2.153590983573798, "grad_norm": 0.11405256390571594, "learning_rate": 2.641284849299569e-05, "loss": 0.0222, "step": 107770 }, { "epoch": 2.1537908157147996, "grad_norm": 0.08685941249132156, "learning_rate": 2.6403736971896366e-05, "loss": 0.0299, "step": 107780 }, { "epoch": 2.1539906478558013, "grad_norm": 0.11430506408214569, "learning_rate": 2.6394626458786025e-05, "loss": 0.0235, "step": 107790 }, { "epoch": 2.1541904799968026, "grad_norm": 0.14271335303783417, "learning_rate": 2.638551695405388e-05, "loss": 0.0255, "step": 107800 }, { "epoch": 2.1543903121378043, "grad_norm": 0.10068986564874649, "learning_rate": 2.6376408458089074e-05, "loss": 0.0309, "step": 107810 }, { "epoch": 2.1545901442788056, "grad_norm": 0.146287739276886, "learning_rate": 2.6367300971280694e-05, "loss": 0.0198, "step": 107820 }, { "epoch": 2.1547899764198073, "grad_norm": 0.10834825783967972, "learning_rate": 2.6358194494017806e-05, "loss": 0.0263, "step": 107830 }, { "epoch": 2.154989808560809, "grad_norm": 0.11942226439714432, "learning_rate": 2.6349089026689423e-05, "loss": 0.0289, "step": 107840 }, { "epoch": 2.1551896407018103, "grad_norm": 0.07966414093971252, "learning_rate": 2.6339984569684523e-05, "loss": 0.0199, "step": 107850 }, { "epoch": 2.155389472842812, "grad_norm": 0.12372744083404541, "learning_rate": 2.6330881123391976e-05, "loss": 0.0262, "step": 107860 }, { "epoch": 2.1555893049838137, "grad_norm": 0.10690166801214218, "learning_rate": 2.6321778688200726e-05, "loss": 0.0241, "step": 107870 }, { "epoch": 2.155789137124815, "grad_norm": 0.27827152609825134, "learning_rate": 2.6312677264499617e-05, "loss": 0.0205, "step": 107880 }, { "epoch": 2.1559889692658167, "grad_norm": 0.11490099877119064, "learning_rate": 2.6303576852677393e-05, "loss": 0.0264, "step": 107890 }, { "epoch": 2.1561888014068185, "grad_norm": 0.10380994528532028, "learning_rate": 2.6294477453122824e-05, "loss": 0.0242, "step": 107900 }, { "epoch": 2.1563886335478197, "grad_norm": 0.11179453879594803, "learning_rate": 2.6285379066224658e-05, "loss": 0.0345, "step": 107910 }, { "epoch": 2.1565884656888215, "grad_norm": 0.09707256406545639, "learning_rate": 2.6276281692371517e-05, "loss": 0.0341, "step": 107920 }, { "epoch": 2.1567882978298227, "grad_norm": 0.11180256307125092, "learning_rate": 2.626718533195205e-05, "loss": 0.0208, "step": 107930 }, { "epoch": 2.1569881299708245, "grad_norm": 0.13180303573608398, "learning_rate": 2.6258089985354815e-05, "loss": 0.0263, "step": 107940 }, { "epoch": 2.157187962111826, "grad_norm": 0.10487353056669235, "learning_rate": 2.6248995652968367e-05, "loss": 0.0256, "step": 107950 }, { "epoch": 2.1573877942528275, "grad_norm": 0.16056309640407562, "learning_rate": 2.6239902335181188e-05, "loss": 0.0363, "step": 107960 }, { "epoch": 2.157587626393829, "grad_norm": 0.13008882105350494, "learning_rate": 2.6230810032381738e-05, "loss": 0.0274, "step": 107970 }, { "epoch": 2.157787458534831, "grad_norm": 0.14726445078849792, "learning_rate": 2.6221718744958436e-05, "loss": 0.0306, "step": 107980 }, { "epoch": 2.157987290675832, "grad_norm": 0.1293565183877945, "learning_rate": 2.6212628473299582e-05, "loss": 0.0242, "step": 107990 }, { "epoch": 2.158187122816834, "grad_norm": 0.11175042390823364, "learning_rate": 2.6203539217793564e-05, "loss": 0.0252, "step": 108000 }, { "epoch": 2.1583869549578356, "grad_norm": 0.10501254349946976, "learning_rate": 2.6194450978828656e-05, "loss": 0.0213, "step": 108010 }, { "epoch": 2.158586787098837, "grad_norm": 0.10519836097955704, "learning_rate": 2.6185363756793053e-05, "loss": 0.0204, "step": 108020 }, { "epoch": 2.1587866192398386, "grad_norm": 0.13110043108463287, "learning_rate": 2.6176277552074956e-05, "loss": 0.0267, "step": 108030 }, { "epoch": 2.15898645138084, "grad_norm": 0.12905441224575043, "learning_rate": 2.6167192365062497e-05, "loss": 0.0244, "step": 108040 }, { "epoch": 2.1591862835218416, "grad_norm": 0.11530957370996475, "learning_rate": 2.6158108196143843e-05, "loss": 0.0213, "step": 108050 }, { "epoch": 2.1593861156628433, "grad_norm": 0.11843691021203995, "learning_rate": 2.6149025045706987e-05, "loss": 0.0313, "step": 108060 }, { "epoch": 2.1595859478038446, "grad_norm": 0.1355869621038437, "learning_rate": 2.6139942914139958e-05, "loss": 0.0262, "step": 108070 }, { "epoch": 2.1597857799448463, "grad_norm": 0.12343554943799973, "learning_rate": 2.613086180183074e-05, "loss": 0.0254, "step": 108080 }, { "epoch": 2.159985612085848, "grad_norm": 0.11509709060192108, "learning_rate": 2.612178170916726e-05, "loss": 0.0339, "step": 108090 }, { "epoch": 2.1601854442268493, "grad_norm": 0.11920137703418732, "learning_rate": 2.61127026365374e-05, "loss": 0.0221, "step": 108100 }, { "epoch": 2.160385276367851, "grad_norm": 0.13495177030563354, "learning_rate": 2.6103624584329e-05, "loss": 0.0273, "step": 108110 }, { "epoch": 2.1605851085088528, "grad_norm": 0.12402407824993134, "learning_rate": 2.609454755292987e-05, "loss": 0.0206, "step": 108120 }, { "epoch": 2.160784940649854, "grad_norm": 0.18543174862861633, "learning_rate": 2.6085471542727725e-05, "loss": 0.0317, "step": 108130 }, { "epoch": 2.1609847727908558, "grad_norm": 0.10819809883832932, "learning_rate": 2.6076396554110315e-05, "loss": 0.0255, "step": 108140 }, { "epoch": 2.161184604931857, "grad_norm": 0.1362621784210205, "learning_rate": 2.6067322587465315e-05, "loss": 0.0206, "step": 108150 }, { "epoch": 2.1613844370728588, "grad_norm": 0.14102977514266968, "learning_rate": 2.6058249643180314e-05, "loss": 0.0265, "step": 108160 }, { "epoch": 2.1615842692138605, "grad_norm": 0.0971793532371521, "learning_rate": 2.604917772164288e-05, "loss": 0.0266, "step": 108170 }, { "epoch": 2.1617841013548618, "grad_norm": 0.1082051619887352, "learning_rate": 2.6040106823240623e-05, "loss": 0.0231, "step": 108180 }, { "epoch": 2.1619839334958635, "grad_norm": 0.12025653570890427, "learning_rate": 2.6031036948360958e-05, "loss": 0.0312, "step": 108190 }, { "epoch": 2.162183765636865, "grad_norm": 0.13304512202739716, "learning_rate": 2.602196809739137e-05, "loss": 0.0251, "step": 108200 }, { "epoch": 2.1623835977778665, "grad_norm": 0.13158152997493744, "learning_rate": 2.6012900270719255e-05, "loss": 0.0274, "step": 108210 }, { "epoch": 2.162583429918868, "grad_norm": 0.13024933636188507, "learning_rate": 2.6003833468731963e-05, "loss": 0.0266, "step": 108220 }, { "epoch": 2.16278326205987, "grad_norm": 0.10397408157587051, "learning_rate": 2.5994767691816833e-05, "loss": 0.0244, "step": 108230 }, { "epoch": 2.162983094200871, "grad_norm": 0.14878253638744354, "learning_rate": 2.5985702940361128e-05, "loss": 0.0305, "step": 108240 }, { "epoch": 2.163182926341873, "grad_norm": 0.12079525738954544, "learning_rate": 2.5976639214752096e-05, "loss": 0.0356, "step": 108250 }, { "epoch": 2.163382758482874, "grad_norm": 0.19924907386302948, "learning_rate": 2.5967576515376868e-05, "loss": 0.0276, "step": 108260 }, { "epoch": 2.163582590623876, "grad_norm": 0.11958178132772446, "learning_rate": 2.595851484262264e-05, "loss": 0.0314, "step": 108270 }, { "epoch": 2.1637824227648776, "grad_norm": 0.11812704801559448, "learning_rate": 2.5949454196876512e-05, "loss": 0.0294, "step": 108280 }, { "epoch": 2.163982254905879, "grad_norm": 0.13326594233512878, "learning_rate": 2.5940394578525495e-05, "loss": 0.0258, "step": 108290 }, { "epoch": 2.1641820870468806, "grad_norm": 0.11643090099096298, "learning_rate": 2.5931335987956623e-05, "loss": 0.0305, "step": 108300 }, { "epoch": 2.1643819191878824, "grad_norm": 0.10503983497619629, "learning_rate": 2.5922278425556868e-05, "loss": 0.0352, "step": 108310 }, { "epoch": 2.1645817513288836, "grad_norm": 0.12100645899772644, "learning_rate": 2.5913221891713134e-05, "loss": 0.0229, "step": 108320 }, { "epoch": 2.1647815834698854, "grad_norm": 0.11047608405351639, "learning_rate": 2.5904166386812322e-05, "loss": 0.0237, "step": 108330 }, { "epoch": 2.164981415610887, "grad_norm": 0.14441503584384918, "learning_rate": 2.5895111911241255e-05, "loss": 0.0222, "step": 108340 }, { "epoch": 2.1651812477518884, "grad_norm": 0.13926498591899872, "learning_rate": 2.588605846538673e-05, "loss": 0.0236, "step": 108350 }, { "epoch": 2.16538107989289, "grad_norm": 0.1420227289199829, "learning_rate": 2.5877006049635483e-05, "loss": 0.0262, "step": 108360 }, { "epoch": 2.1655809120338914, "grad_norm": 0.10736168920993805, "learning_rate": 2.5867954664374224e-05, "loss": 0.0292, "step": 108370 }, { "epoch": 2.165780744174893, "grad_norm": 0.11168482154607773, "learning_rate": 2.585890430998963e-05, "loss": 0.0241, "step": 108380 }, { "epoch": 2.165980576315895, "grad_norm": 0.14833161234855652, "learning_rate": 2.584985498686827e-05, "loss": 0.0268, "step": 108390 }, { "epoch": 2.166180408456896, "grad_norm": 0.10595346987247467, "learning_rate": 2.5840806695396724e-05, "loss": 0.0252, "step": 108400 }, { "epoch": 2.166380240597898, "grad_norm": 0.12885832786560059, "learning_rate": 2.5831759435961576e-05, "loss": 0.0208, "step": 108410 }, { "epoch": 2.1665800727388995, "grad_norm": 0.08171884715557098, "learning_rate": 2.5822713208949247e-05, "loss": 0.0274, "step": 108420 }, { "epoch": 2.166779904879901, "grad_norm": 0.13323237001895905, "learning_rate": 2.581366801474619e-05, "loss": 0.0267, "step": 108430 }, { "epoch": 2.1669797370209025, "grad_norm": 0.12224569171667099, "learning_rate": 2.5804623853738785e-05, "loss": 0.027, "step": 108440 }, { "epoch": 2.167179569161904, "grad_norm": 0.12637513875961304, "learning_rate": 2.579558072631344e-05, "loss": 0.03, "step": 108450 }, { "epoch": 2.1673794013029055, "grad_norm": 0.10254169255495071, "learning_rate": 2.5786538632856406e-05, "loss": 0.0278, "step": 108460 }, { "epoch": 2.1675792334439072, "grad_norm": 0.13198308646678925, "learning_rate": 2.5777497573753956e-05, "loss": 0.0266, "step": 108470 }, { "epoch": 2.1677790655849085, "grad_norm": 0.10952018201351166, "learning_rate": 2.576845754939231e-05, "loss": 0.0176, "step": 108480 }, { "epoch": 2.1679788977259102, "grad_norm": 0.1210712417960167, "learning_rate": 2.5759418560157645e-05, "loss": 0.0375, "step": 108490 }, { "epoch": 2.168178729866912, "grad_norm": 0.1432676613330841, "learning_rate": 2.5750380606436087e-05, "loss": 0.0263, "step": 108500 }, { "epoch": 2.1683785620079132, "grad_norm": 0.09739325195550919, "learning_rate": 2.574134368861373e-05, "loss": 0.0236, "step": 108510 }, { "epoch": 2.168578394148915, "grad_norm": 0.14809612929821014, "learning_rate": 2.573230780707661e-05, "loss": 0.0257, "step": 108520 }, { "epoch": 2.1687782262899167, "grad_norm": 0.14495129883289337, "learning_rate": 2.5723272962210687e-05, "loss": 0.0255, "step": 108530 }, { "epoch": 2.168978058430918, "grad_norm": 0.0869595855474472, "learning_rate": 2.571423915440197e-05, "loss": 0.0223, "step": 108540 }, { "epoch": 2.1691778905719197, "grad_norm": 0.11587858200073242, "learning_rate": 2.5705206384036358e-05, "loss": 0.0233, "step": 108550 }, { "epoch": 2.169377722712921, "grad_norm": 0.11748606711626053, "learning_rate": 2.5696174651499676e-05, "loss": 0.0247, "step": 108560 }, { "epoch": 2.1695775548539227, "grad_norm": 0.113408163189888, "learning_rate": 2.5687143957177772e-05, "loss": 0.0265, "step": 108570 }, { "epoch": 2.1697773869949244, "grad_norm": 0.13066521286964417, "learning_rate": 2.5678114301456406e-05, "loss": 0.019, "step": 108580 }, { "epoch": 2.1699772191359257, "grad_norm": 0.13000419735908508, "learning_rate": 2.5669085684721316e-05, "loss": 0.0193, "step": 108590 }, { "epoch": 2.1701770512769274, "grad_norm": 0.11818055808544159, "learning_rate": 2.5660058107358197e-05, "loss": 0.0308, "step": 108600 }, { "epoch": 2.170376883417929, "grad_norm": 0.11286205798387527, "learning_rate": 2.5651031569752677e-05, "loss": 0.0214, "step": 108610 }, { "epoch": 2.1705767155589304, "grad_norm": 0.10707753896713257, "learning_rate": 2.5642006072290358e-05, "loss": 0.0238, "step": 108620 }, { "epoch": 2.170776547699932, "grad_norm": 0.15411387383937836, "learning_rate": 2.563298161535679e-05, "loss": 0.0216, "step": 108630 }, { "epoch": 2.1709763798409334, "grad_norm": 0.10690709948539734, "learning_rate": 2.5623958199337484e-05, "loss": 0.0217, "step": 108640 }, { "epoch": 2.171176211981935, "grad_norm": 0.15947850048542023, "learning_rate": 2.5614935824617925e-05, "loss": 0.0271, "step": 108650 }, { "epoch": 2.171376044122937, "grad_norm": 0.13131089508533478, "learning_rate": 2.5605914491583492e-05, "loss": 0.0201, "step": 108660 }, { "epoch": 2.171575876263938, "grad_norm": 0.11811499297618866, "learning_rate": 2.559689420061956e-05, "loss": 0.0319, "step": 108670 }, { "epoch": 2.17177570840494, "grad_norm": 0.09066084772348404, "learning_rate": 2.5587874952111513e-05, "loss": 0.0271, "step": 108680 }, { "epoch": 2.1719755405459416, "grad_norm": 0.11155451089143753, "learning_rate": 2.557885674644459e-05, "loss": 0.0294, "step": 108690 }, { "epoch": 2.172175372686943, "grad_norm": 0.15186984837055206, "learning_rate": 2.5569839584004034e-05, "loss": 0.0254, "step": 108700 }, { "epoch": 2.1723752048279446, "grad_norm": 0.10440433770418167, "learning_rate": 2.5560823465175055e-05, "loss": 0.0265, "step": 108710 }, { "epoch": 2.1725750369689463, "grad_norm": 0.11767846345901489, "learning_rate": 2.55518083903428e-05, "loss": 0.0314, "step": 108720 }, { "epoch": 2.1727748691099475, "grad_norm": 0.09554074704647064, "learning_rate": 2.5542794359892386e-05, "loss": 0.0261, "step": 108730 }, { "epoch": 2.1729747012509493, "grad_norm": 0.10242623835802078, "learning_rate": 2.5533781374208854e-05, "loss": 0.0283, "step": 108740 }, { "epoch": 2.1731745333919505, "grad_norm": 0.11638106405735016, "learning_rate": 2.5524769433677266e-05, "loss": 0.0258, "step": 108750 }, { "epoch": 2.1733743655329523, "grad_norm": 0.0944647565484047, "learning_rate": 2.551575853868252e-05, "loss": 0.0249, "step": 108760 }, { "epoch": 2.173574197673954, "grad_norm": 0.1335904598236084, "learning_rate": 2.5506748689609616e-05, "loss": 0.0225, "step": 108770 }, { "epoch": 2.1737740298149553, "grad_norm": 0.13397003710269928, "learning_rate": 2.549773988684343e-05, "loss": 0.0316, "step": 108780 }, { "epoch": 2.173973861955957, "grad_norm": 0.15448814630508423, "learning_rate": 2.548873213076876e-05, "loss": 0.0219, "step": 108790 }, { "epoch": 2.1741736940969587, "grad_norm": 0.14174365997314453, "learning_rate": 2.5479725421770408e-05, "loss": 0.0291, "step": 108800 }, { "epoch": 2.17437352623796, "grad_norm": 0.10606777667999268, "learning_rate": 2.547071976023318e-05, "loss": 0.0287, "step": 108810 }, { "epoch": 2.1745733583789617, "grad_norm": 0.1284995824098587, "learning_rate": 2.546171514654172e-05, "loss": 0.0244, "step": 108820 }, { "epoch": 2.1747731905199634, "grad_norm": 0.09283427894115448, "learning_rate": 2.5452711581080714e-05, "loss": 0.0288, "step": 108830 }, { "epoch": 2.1749730226609647, "grad_norm": 0.13999563455581665, "learning_rate": 2.5443709064234757e-05, "loss": 0.0289, "step": 108840 }, { "epoch": 2.1751728548019664, "grad_norm": 0.11586388945579529, "learning_rate": 2.5434707596388473e-05, "loss": 0.0252, "step": 108850 }, { "epoch": 2.1753726869429677, "grad_norm": 0.1367255449295044, "learning_rate": 2.5425707177926333e-05, "loss": 0.0265, "step": 108860 }, { "epoch": 2.1755725190839694, "grad_norm": 0.13415509462356567, "learning_rate": 2.5416707809232844e-05, "loss": 0.0262, "step": 108870 }, { "epoch": 2.175772351224971, "grad_norm": 0.1389976292848587, "learning_rate": 2.5407709490692423e-05, "loss": 0.0278, "step": 108880 }, { "epoch": 2.1759721833659724, "grad_norm": 0.12022534757852554, "learning_rate": 2.5398712222689486e-05, "loss": 0.018, "step": 108890 }, { "epoch": 2.176172015506974, "grad_norm": 0.1247599646449089, "learning_rate": 2.5389716005608367e-05, "loss": 0.0276, "step": 108900 }, { "epoch": 2.176371847647976, "grad_norm": 0.1233225017786026, "learning_rate": 2.5380720839833368e-05, "loss": 0.0241, "step": 108910 }, { "epoch": 2.176571679788977, "grad_norm": 0.11738444119691849, "learning_rate": 2.5371726725748767e-05, "loss": 0.0237, "step": 108920 }, { "epoch": 2.176771511929979, "grad_norm": 0.17197440564632416, "learning_rate": 2.5362733663738718e-05, "loss": 0.0275, "step": 108930 }, { "epoch": 2.1769713440709806, "grad_norm": 0.1520610898733139, "learning_rate": 2.5353741654187445e-05, "loss": 0.0276, "step": 108940 }, { "epoch": 2.177171176211982, "grad_norm": 0.14466360211372375, "learning_rate": 2.5344750697479074e-05, "loss": 0.0271, "step": 108950 }, { "epoch": 2.1773710083529836, "grad_norm": 0.14801405370235443, "learning_rate": 2.5335760793997642e-05, "loss": 0.0199, "step": 108960 }, { "epoch": 2.177570840493985, "grad_norm": 0.146243616938591, "learning_rate": 2.5326771944127192e-05, "loss": 0.0287, "step": 108970 }, { "epoch": 2.1777706726349866, "grad_norm": 0.14958402514457703, "learning_rate": 2.5317784148251728e-05, "loss": 0.0269, "step": 108980 }, { "epoch": 2.1779705047759883, "grad_norm": 0.17825573682785034, "learning_rate": 2.530879740675518e-05, "loss": 0.0227, "step": 108990 }, { "epoch": 2.1781703369169896, "grad_norm": 0.13412706553936005, "learning_rate": 2.5299811720021442e-05, "loss": 0.0374, "step": 109000 }, { "epoch": 2.1783701690579913, "grad_norm": 0.11815162003040314, "learning_rate": 2.5290827088434376e-05, "loss": 0.0238, "step": 109010 }, { "epoch": 2.178570001198993, "grad_norm": 0.11035597324371338, "learning_rate": 2.5281843512377785e-05, "loss": 0.0336, "step": 109020 }, { "epoch": 2.1787698333399943, "grad_norm": 0.12460391223430634, "learning_rate": 2.5272860992235425e-05, "loss": 0.0283, "step": 109030 }, { "epoch": 2.178969665480996, "grad_norm": 0.11123137921094894, "learning_rate": 2.526387952839102e-05, "loss": 0.0208, "step": 109040 }, { "epoch": 2.1791694976219977, "grad_norm": 0.12793828547000885, "learning_rate": 2.5254899121228272e-05, "loss": 0.0314, "step": 109050 }, { "epoch": 2.179369329762999, "grad_norm": 0.10194752365350723, "learning_rate": 2.5245919771130745e-05, "loss": 0.0204, "step": 109060 }, { "epoch": 2.1795691619040007, "grad_norm": 0.08479057252407074, "learning_rate": 2.5236941478482025e-05, "loss": 0.0244, "step": 109070 }, { "epoch": 2.179768994045002, "grad_norm": 0.10833411663770676, "learning_rate": 2.5227964243665725e-05, "loss": 0.0196, "step": 109080 }, { "epoch": 2.1799688261860037, "grad_norm": 0.09023620933294296, "learning_rate": 2.521898806706525e-05, "loss": 0.0304, "step": 109090 }, { "epoch": 2.1801686583270055, "grad_norm": 0.11653943359851837, "learning_rate": 2.52100129490641e-05, "loss": 0.0227, "step": 109100 }, { "epoch": 2.1803684904680067, "grad_norm": 0.12579220533370972, "learning_rate": 2.520103889004564e-05, "loss": 0.0243, "step": 109110 }, { "epoch": 2.1805683226090085, "grad_norm": 0.17478692531585693, "learning_rate": 2.519206589039325e-05, "loss": 0.0323, "step": 109120 }, { "epoch": 2.18076815475001, "grad_norm": 0.1324758678674698, "learning_rate": 2.5183093950490233e-05, "loss": 0.0225, "step": 109130 }, { "epoch": 2.1809679868910115, "grad_norm": 0.1320914328098297, "learning_rate": 2.5174123070719848e-05, "loss": 0.0312, "step": 109140 }, { "epoch": 2.181167819032013, "grad_norm": 0.11290293186903, "learning_rate": 2.5165153251465334e-05, "loss": 0.0282, "step": 109150 }, { "epoch": 2.181367651173015, "grad_norm": 0.15468911826610565, "learning_rate": 2.515618449310982e-05, "loss": 0.0222, "step": 109160 }, { "epoch": 2.181567483314016, "grad_norm": 0.09223183244466782, "learning_rate": 2.5147216796036477e-05, "loss": 0.0335, "step": 109170 }, { "epoch": 2.181767315455018, "grad_norm": 0.10619982331991196, "learning_rate": 2.5138250160628407e-05, "loss": 0.0239, "step": 109180 }, { "epoch": 2.181967147596019, "grad_norm": 0.1321752369403839, "learning_rate": 2.5129284587268586e-05, "loss": 0.0301, "step": 109190 }, { "epoch": 2.182166979737021, "grad_norm": 0.14746087789535522, "learning_rate": 2.5120320076340026e-05, "loss": 0.0221, "step": 109200 }, { "epoch": 2.1823668118780226, "grad_norm": 0.08657997846603394, "learning_rate": 2.5111356628225703e-05, "loss": 0.0185, "step": 109210 }, { "epoch": 2.182566644019024, "grad_norm": 0.12587608397006989, "learning_rate": 2.5102394243308526e-05, "loss": 0.0287, "step": 109220 }, { "epoch": 2.1827664761600256, "grad_norm": 0.1300879865884781, "learning_rate": 2.5093432921971304e-05, "loss": 0.0304, "step": 109230 }, { "epoch": 2.1829663083010273, "grad_norm": 0.10792338103055954, "learning_rate": 2.5084472664596874e-05, "loss": 0.0239, "step": 109240 }, { "epoch": 2.1831661404420286, "grad_norm": 0.12101587653160095, "learning_rate": 2.5075513471568002e-05, "loss": 0.0234, "step": 109250 }, { "epoch": 2.1833659725830303, "grad_norm": 0.11268139630556107, "learning_rate": 2.5066555343267405e-05, "loss": 0.0285, "step": 109260 }, { "epoch": 2.1835658047240316, "grad_norm": 0.09600891172885895, "learning_rate": 2.5057598280077766e-05, "loss": 0.0295, "step": 109270 }, { "epoch": 2.1837656368650333, "grad_norm": 0.11636912077665329, "learning_rate": 2.50486422823817e-05, "loss": 0.0284, "step": 109280 }, { "epoch": 2.183965469006035, "grad_norm": 0.10584054887294769, "learning_rate": 2.50396873505618e-05, "loss": 0.0234, "step": 109290 }, { "epoch": 2.1841653011470363, "grad_norm": 0.13127462565898895, "learning_rate": 2.5030733485000607e-05, "loss": 0.0231, "step": 109300 }, { "epoch": 2.184365133288038, "grad_norm": 0.13970640301704407, "learning_rate": 2.502178068608061e-05, "loss": 0.0317, "step": 109310 }, { "epoch": 2.1845649654290398, "grad_norm": 0.1314551681280136, "learning_rate": 2.501282895418427e-05, "loss": 0.024, "step": 109320 }, { "epoch": 2.184764797570041, "grad_norm": 0.09698041528463364, "learning_rate": 2.5003878289693955e-05, "loss": 0.0224, "step": 109330 }, { "epoch": 2.1849646297110428, "grad_norm": 0.11406995356082916, "learning_rate": 2.4994928692992027e-05, "loss": 0.0263, "step": 109340 }, { "epoch": 2.185164461852044, "grad_norm": 0.23767074942588806, "learning_rate": 2.4985980164460838e-05, "loss": 0.0381, "step": 109350 }, { "epoch": 2.1853642939930458, "grad_norm": 0.1501789391040802, "learning_rate": 2.497703270448261e-05, "loss": 0.0269, "step": 109360 }, { "epoch": 2.1855641261340475, "grad_norm": 0.11603754758834839, "learning_rate": 2.4968086313439576e-05, "loss": 0.0241, "step": 109370 }, { "epoch": 2.1857639582750488, "grad_norm": 0.1534869521856308, "learning_rate": 2.4959140991713916e-05, "loss": 0.0232, "step": 109380 }, { "epoch": 2.1859637904160505, "grad_norm": 0.09245052188634872, "learning_rate": 2.4950196739687743e-05, "loss": 0.0306, "step": 109390 }, { "epoch": 2.186163622557052, "grad_norm": 0.12254449725151062, "learning_rate": 2.4941253557743148e-05, "loss": 0.0234, "step": 109400 }, { "epoch": 2.1863634546980535, "grad_norm": 0.11944126337766647, "learning_rate": 2.493231144626217e-05, "loss": 0.0234, "step": 109410 }, { "epoch": 2.186563286839055, "grad_norm": 0.12394709885120392, "learning_rate": 2.492337040562681e-05, "loss": 0.0167, "step": 109420 }, { "epoch": 2.186763118980057, "grad_norm": 0.12106998264789581, "learning_rate": 2.4914430436218968e-05, "loss": 0.0223, "step": 109430 }, { "epoch": 2.186962951121058, "grad_norm": 0.09810896217823029, "learning_rate": 2.490549153842059e-05, "loss": 0.0347, "step": 109440 }, { "epoch": 2.18716278326206, "grad_norm": 0.1564602553844452, "learning_rate": 2.489655371261353e-05, "loss": 0.028, "step": 109450 }, { "epoch": 2.187362615403061, "grad_norm": 0.13371510803699493, "learning_rate": 2.4887616959179566e-05, "loss": 0.0256, "step": 109460 }, { "epoch": 2.187562447544063, "grad_norm": 0.10373780876398087, "learning_rate": 2.487868127850045e-05, "loss": 0.0239, "step": 109470 }, { "epoch": 2.1877622796850646, "grad_norm": 0.16330046951770782, "learning_rate": 2.4869746670957955e-05, "loss": 0.0204, "step": 109480 }, { "epoch": 2.187962111826066, "grad_norm": 0.10674109309911728, "learning_rate": 2.4860813136933704e-05, "loss": 0.0215, "step": 109490 }, { "epoch": 2.1881619439670676, "grad_norm": 0.11632965505123138, "learning_rate": 2.485188067680933e-05, "loss": 0.0325, "step": 109500 }, { "epoch": 2.1883617761080694, "grad_norm": 0.0904250293970108, "learning_rate": 2.4842949290966412e-05, "loss": 0.0199, "step": 109510 }, { "epoch": 2.1885616082490706, "grad_norm": 0.09706079959869385, "learning_rate": 2.483401897978649e-05, "loss": 0.0249, "step": 109520 }, { "epoch": 2.1887614403900724, "grad_norm": 0.10329370200634003, "learning_rate": 2.482508974365104e-05, "loss": 0.0218, "step": 109530 }, { "epoch": 2.188961272531074, "grad_norm": 0.10726168006658554, "learning_rate": 2.4816161582941516e-05, "loss": 0.0243, "step": 109540 }, { "epoch": 2.1891611046720754, "grad_norm": 0.12289921194314957, "learning_rate": 2.480723449803932e-05, "loss": 0.021, "step": 109550 }, { "epoch": 2.189360936813077, "grad_norm": 0.11295690387487411, "learning_rate": 2.479830848932575e-05, "loss": 0.0188, "step": 109560 }, { "epoch": 2.1895607689540784, "grad_norm": 0.09277563542127609, "learning_rate": 2.478938355718216e-05, "loss": 0.0202, "step": 109570 }, { "epoch": 2.18976060109508, "grad_norm": 0.10296061635017395, "learning_rate": 2.4780459701989813e-05, "loss": 0.0269, "step": 109580 }, { "epoch": 2.189960433236082, "grad_norm": 0.10835998505353928, "learning_rate": 2.477153692412988e-05, "loss": 0.0244, "step": 109590 }, { "epoch": 2.190160265377083, "grad_norm": 0.13601337373256683, "learning_rate": 2.4762615223983538e-05, "loss": 0.0329, "step": 109600 }, { "epoch": 2.190360097518085, "grad_norm": 0.13765762746334076, "learning_rate": 2.4753694601931892e-05, "loss": 0.0253, "step": 109610 }, { "epoch": 2.1905599296590865, "grad_norm": 0.13907738029956818, "learning_rate": 2.474477505835608e-05, "loss": 0.0296, "step": 109620 }, { "epoch": 2.190759761800088, "grad_norm": 0.08517351746559143, "learning_rate": 2.473585659363706e-05, "loss": 0.0233, "step": 109630 }, { "epoch": 2.1909595939410895, "grad_norm": 0.12084309011697769, "learning_rate": 2.472693920815583e-05, "loss": 0.0256, "step": 109640 }, { "epoch": 2.1911594260820912, "grad_norm": 0.14177486300468445, "learning_rate": 2.4718022902293332e-05, "loss": 0.0289, "step": 109650 }, { "epoch": 2.1913592582230925, "grad_norm": 0.09587755799293518, "learning_rate": 2.470910767643045e-05, "loss": 0.0214, "step": 109660 }, { "epoch": 2.1915590903640942, "grad_norm": 0.1580742597579956, "learning_rate": 2.4700193530948022e-05, "loss": 0.0239, "step": 109670 }, { "epoch": 2.1917589225050955, "grad_norm": 0.10167571157217026, "learning_rate": 2.4691280466226857e-05, "loss": 0.0224, "step": 109680 }, { "epoch": 2.1919587546460972, "grad_norm": 0.15732896327972412, "learning_rate": 2.4682368482647717e-05, "loss": 0.0269, "step": 109690 }, { "epoch": 2.192158586787099, "grad_norm": 0.09414095431566238, "learning_rate": 2.4673457580591242e-05, "loss": 0.0279, "step": 109700 }, { "epoch": 2.1923584189281002, "grad_norm": 0.13306184113025665, "learning_rate": 2.4664547760438156e-05, "loss": 0.0187, "step": 109710 }, { "epoch": 2.192558251069102, "grad_norm": 0.11294449120759964, "learning_rate": 2.465563902256906e-05, "loss": 0.0326, "step": 109720 }, { "epoch": 2.1927580832101037, "grad_norm": 0.14812196791172028, "learning_rate": 2.4646731367364495e-05, "loss": 0.0272, "step": 109730 }, { "epoch": 2.192957915351105, "grad_norm": 0.10070531070232391, "learning_rate": 2.4637824795204965e-05, "loss": 0.0285, "step": 109740 }, { "epoch": 2.1931577474921067, "grad_norm": 0.11864849925041199, "learning_rate": 2.462891930647101e-05, "loss": 0.0277, "step": 109750 }, { "epoch": 2.1933575796331084, "grad_norm": 0.1344744712114334, "learning_rate": 2.4620014901542993e-05, "loss": 0.0257, "step": 109760 }, { "epoch": 2.1935574117741097, "grad_norm": 0.16611507534980774, "learning_rate": 2.4611111580801316e-05, "loss": 0.0363, "step": 109770 }, { "epoch": 2.1937572439151114, "grad_norm": 0.14794978499412537, "learning_rate": 2.460220934462631e-05, "loss": 0.0345, "step": 109780 }, { "epoch": 2.1939570760561127, "grad_norm": 0.12276281416416168, "learning_rate": 2.459330819339826e-05, "loss": 0.0275, "step": 109790 }, { "epoch": 2.1941569081971144, "grad_norm": 0.09926015138626099, "learning_rate": 2.4584408127497415e-05, "loss": 0.0217, "step": 109800 }, { "epoch": 2.194356740338116, "grad_norm": 0.16204452514648438, "learning_rate": 2.4575509147303966e-05, "loss": 0.0284, "step": 109810 }, { "epoch": 2.1945565724791174, "grad_norm": 0.12339076399803162, "learning_rate": 2.456661125319808e-05, "loss": 0.0293, "step": 109820 }, { "epoch": 2.194756404620119, "grad_norm": 0.12625554203987122, "learning_rate": 2.4557714445559793e-05, "loss": 0.0345, "step": 109830 }, { "epoch": 2.194956236761121, "grad_norm": 0.107312873005867, "learning_rate": 2.4548818724769234e-05, "loss": 0.0307, "step": 109840 }, { "epoch": 2.195156068902122, "grad_norm": 0.11726992577314377, "learning_rate": 2.4539924091206395e-05, "loss": 0.0327, "step": 109850 }, { "epoch": 2.195355901043124, "grad_norm": 0.14049285650253296, "learning_rate": 2.4531030545251217e-05, "loss": 0.0176, "step": 109860 }, { "epoch": 2.1955557331841256, "grad_norm": 0.08742108941078186, "learning_rate": 2.452213808728362e-05, "loss": 0.0239, "step": 109870 }, { "epoch": 2.195755565325127, "grad_norm": 0.09363111108541489, "learning_rate": 2.451324671768348e-05, "loss": 0.0218, "step": 109880 }, { "epoch": 2.1959553974661286, "grad_norm": 0.09916911274194717, "learning_rate": 2.4504356436830617e-05, "loss": 0.0228, "step": 109890 }, { "epoch": 2.19615522960713, "grad_norm": 0.1411137878894806, "learning_rate": 2.4495467245104815e-05, "loss": 0.023, "step": 109900 }, { "epoch": 2.1963550617481316, "grad_norm": 0.1781187206506729, "learning_rate": 2.4486579142885798e-05, "loss": 0.0288, "step": 109910 }, { "epoch": 2.1965548938891333, "grad_norm": 0.12826727330684662, "learning_rate": 2.447769213055325e-05, "loss": 0.0259, "step": 109920 }, { "epoch": 2.1967547260301346, "grad_norm": 0.12049910426139832, "learning_rate": 2.4468806208486804e-05, "loss": 0.0302, "step": 109930 }, { "epoch": 2.1969545581711363, "grad_norm": 0.12350210547447205, "learning_rate": 2.4459921377066058e-05, "loss": 0.0303, "step": 109940 }, { "epoch": 2.197154390312138, "grad_norm": 0.10514543205499649, "learning_rate": 2.445103763667057e-05, "loss": 0.0237, "step": 109950 }, { "epoch": 2.1973542224531393, "grad_norm": 0.10453378409147263, "learning_rate": 2.4442154987679806e-05, "loss": 0.0193, "step": 109960 }, { "epoch": 2.197554054594141, "grad_norm": 0.14806632697582245, "learning_rate": 2.4433273430473204e-05, "loss": 0.0208, "step": 109970 }, { "epoch": 2.1977538867351427, "grad_norm": 0.1846427023410797, "learning_rate": 2.442439296543023e-05, "loss": 0.0255, "step": 109980 }, { "epoch": 2.197953718876144, "grad_norm": 0.10006533563137054, "learning_rate": 2.4415513592930194e-05, "loss": 0.0196, "step": 109990 }, { "epoch": 2.1981535510171457, "grad_norm": 0.09902451932430267, "learning_rate": 2.4406635313352412e-05, "loss": 0.0242, "step": 110000 }, { "epoch": 2.198353383158147, "grad_norm": 0.12449783086776733, "learning_rate": 2.4397758127076132e-05, "loss": 0.0288, "step": 110010 }, { "epoch": 2.1985532152991487, "grad_norm": 0.16461268067359924, "learning_rate": 2.4388882034480638e-05, "loss": 0.0301, "step": 110020 }, { "epoch": 2.1987530474401504, "grad_norm": 0.18370600044727325, "learning_rate": 2.4380007035945025e-05, "loss": 0.0277, "step": 110030 }, { "epoch": 2.1989528795811517, "grad_norm": 0.12525756657123566, "learning_rate": 2.4371133131848452e-05, "loss": 0.0216, "step": 110040 }, { "epoch": 2.1991527117221534, "grad_norm": 0.12070732563734055, "learning_rate": 2.4362260322569987e-05, "loss": 0.0262, "step": 110050 }, { "epoch": 2.199352543863155, "grad_norm": 0.12508463859558105, "learning_rate": 2.4353388608488666e-05, "loss": 0.0329, "step": 110060 }, { "epoch": 2.1995523760041564, "grad_norm": 0.14403773844242096, "learning_rate": 2.4344517989983466e-05, "loss": 0.0321, "step": 110070 }, { "epoch": 2.199752208145158, "grad_norm": 0.1512337177991867, "learning_rate": 2.4335648467433332e-05, "loss": 0.0323, "step": 110080 }, { "epoch": 2.1999520402861594, "grad_norm": 0.13853254914283752, "learning_rate": 2.4326780041217163e-05, "loss": 0.0197, "step": 110090 }, { "epoch": 2.200151872427161, "grad_norm": 0.09142933040857315, "learning_rate": 2.4317912711713748e-05, "loss": 0.0245, "step": 110100 }, { "epoch": 2.200351704568163, "grad_norm": 0.1022474467754364, "learning_rate": 2.4309046479301945e-05, "loss": 0.0211, "step": 110110 }, { "epoch": 2.200551536709164, "grad_norm": 0.14370085299015045, "learning_rate": 2.4300181344360502e-05, "loss": 0.0253, "step": 110120 }, { "epoch": 2.200751368850166, "grad_norm": 0.13321001827716827, "learning_rate": 2.4291317307268075e-05, "loss": 0.0258, "step": 110130 }, { "epoch": 2.2009512009911676, "grad_norm": 0.11647137254476547, "learning_rate": 2.428245436840333e-05, "loss": 0.0272, "step": 110140 }, { "epoch": 2.201151033132169, "grad_norm": 0.10613443702459335, "learning_rate": 2.4273592528144923e-05, "loss": 0.026, "step": 110150 }, { "epoch": 2.2013508652731706, "grad_norm": 0.1488417536020279, "learning_rate": 2.4264731786871365e-05, "loss": 0.0274, "step": 110160 }, { "epoch": 2.201550697414172, "grad_norm": 0.14845377206802368, "learning_rate": 2.425587214496119e-05, "loss": 0.0267, "step": 110170 }, { "epoch": 2.2017505295551736, "grad_norm": 0.10994817316532135, "learning_rate": 2.4247013602792863e-05, "loss": 0.0228, "step": 110180 }, { "epoch": 2.2019503616961753, "grad_norm": 0.12750783562660217, "learning_rate": 2.42381561607448e-05, "loss": 0.0252, "step": 110190 }, { "epoch": 2.2021501938371766, "grad_norm": 0.12351779639720917, "learning_rate": 2.4229299819195388e-05, "loss": 0.0221, "step": 110200 }, { "epoch": 2.2023500259781783, "grad_norm": 0.14765118062496185, "learning_rate": 2.4220444578522937e-05, "loss": 0.0283, "step": 110210 }, { "epoch": 2.20254985811918, "grad_norm": 0.11077858507633209, "learning_rate": 2.4211590439105762e-05, "loss": 0.026, "step": 110220 }, { "epoch": 2.2027496902601813, "grad_norm": 0.11471887677907944, "learning_rate": 2.4202737401322028e-05, "loss": 0.0359, "step": 110230 }, { "epoch": 2.202949522401183, "grad_norm": 0.1273171752691269, "learning_rate": 2.419388546554998e-05, "loss": 0.0326, "step": 110240 }, { "epoch": 2.2031493545421847, "grad_norm": 0.1353548914194107, "learning_rate": 2.4185034632167752e-05, "loss": 0.0194, "step": 110250 }, { "epoch": 2.203349186683186, "grad_norm": 0.14166732132434845, "learning_rate": 2.4176184901553412e-05, "loss": 0.0335, "step": 110260 }, { "epoch": 2.2035490188241877, "grad_norm": 0.11393711715936661, "learning_rate": 2.4167336274085005e-05, "loss": 0.024, "step": 110270 }, { "epoch": 2.203748850965189, "grad_norm": 0.11252322793006897, "learning_rate": 2.4158488750140533e-05, "loss": 0.0296, "step": 110280 }, { "epoch": 2.2039486831061907, "grad_norm": 0.12852098047733307, "learning_rate": 2.414964233009796e-05, "loss": 0.0174, "step": 110290 }, { "epoch": 2.2041485152471925, "grad_norm": 0.12152126431465149, "learning_rate": 2.414079701433517e-05, "loss": 0.0222, "step": 110300 }, { "epoch": 2.2043483473881937, "grad_norm": 0.16786272823810577, "learning_rate": 2.4131952803230028e-05, "loss": 0.0244, "step": 110310 }, { "epoch": 2.2045481795291955, "grad_norm": 0.1096964105963707, "learning_rate": 2.4123109697160344e-05, "loss": 0.027, "step": 110320 }, { "epoch": 2.204748011670197, "grad_norm": 0.12652677297592163, "learning_rate": 2.4114267696503873e-05, "loss": 0.0242, "step": 110330 }, { "epoch": 2.2049478438111985, "grad_norm": 0.14152012765407562, "learning_rate": 2.4105426801638336e-05, "loss": 0.0255, "step": 110340 }, { "epoch": 2.2051476759522, "grad_norm": 0.09197831153869629, "learning_rate": 2.4096587012941413e-05, "loss": 0.0248, "step": 110350 }, { "epoch": 2.205347508093202, "grad_norm": 0.14134126901626587, "learning_rate": 2.4087748330790687e-05, "loss": 0.0225, "step": 110360 }, { "epoch": 2.205547340234203, "grad_norm": 0.08601880818605423, "learning_rate": 2.4078910755563734e-05, "loss": 0.0253, "step": 110370 }, { "epoch": 2.205747172375205, "grad_norm": 0.1081140786409378, "learning_rate": 2.407007428763813e-05, "loss": 0.0297, "step": 110380 }, { "epoch": 2.205947004516206, "grad_norm": 0.11024714261293411, "learning_rate": 2.4061238927391298e-05, "loss": 0.0268, "step": 110390 }, { "epoch": 2.206146836657208, "grad_norm": 0.11361217498779297, "learning_rate": 2.405240467520069e-05, "loss": 0.0286, "step": 110400 }, { "epoch": 2.2063466687982096, "grad_norm": 0.08050093799829483, "learning_rate": 2.404357153144367e-05, "loss": 0.0253, "step": 110410 }, { "epoch": 2.206546500939211, "grad_norm": 0.1203441321849823, "learning_rate": 2.4034739496497627e-05, "loss": 0.0243, "step": 110420 }, { "epoch": 2.2067463330802126, "grad_norm": 0.14687511324882507, "learning_rate": 2.402590857073979e-05, "loss": 0.0235, "step": 110430 }, { "epoch": 2.2069461652212143, "grad_norm": 0.12602603435516357, "learning_rate": 2.401707875454743e-05, "loss": 0.0255, "step": 110440 }, { "epoch": 2.2071459973622156, "grad_norm": 0.1265842169523239, "learning_rate": 2.4008250048297726e-05, "loss": 0.0214, "step": 110450 }, { "epoch": 2.2073458295032173, "grad_norm": 0.1458713859319687, "learning_rate": 2.3999422452367838e-05, "loss": 0.036, "step": 110460 }, { "epoch": 2.207545661644219, "grad_norm": 0.11955155432224274, "learning_rate": 2.3990595967134853e-05, "loss": 0.0216, "step": 110470 }, { "epoch": 2.2077454937852203, "grad_norm": 0.13922862708568573, "learning_rate": 2.3981770592975832e-05, "loss": 0.033, "step": 110480 }, { "epoch": 2.207945325926222, "grad_norm": 0.17318806052207947, "learning_rate": 2.3972946330267787e-05, "loss": 0.0334, "step": 110490 }, { "epoch": 2.2081451580672233, "grad_norm": 0.1249794289469719, "learning_rate": 2.3964123179387632e-05, "loss": 0.0235, "step": 110500 }, { "epoch": 2.208344990208225, "grad_norm": 0.11902852356433868, "learning_rate": 2.3955301140712315e-05, "loss": 0.0283, "step": 110510 }, { "epoch": 2.2085448223492268, "grad_norm": 0.15431076288223267, "learning_rate": 2.3946480214618715e-05, "loss": 0.0271, "step": 110520 }, { "epoch": 2.208744654490228, "grad_norm": 0.11522480100393295, "learning_rate": 2.39376604014836e-05, "loss": 0.0317, "step": 110530 }, { "epoch": 2.2089444866312298, "grad_norm": 0.1635402888059616, "learning_rate": 2.3928841701683746e-05, "loss": 0.0269, "step": 110540 }, { "epoch": 2.2091443187722315, "grad_norm": 0.11443296074867249, "learning_rate": 2.3920024115595884e-05, "loss": 0.029, "step": 110550 }, { "epoch": 2.2093441509132328, "grad_norm": 0.1194886565208435, "learning_rate": 2.3911207643596685e-05, "loss": 0.0263, "step": 110560 }, { "epoch": 2.2095439830542345, "grad_norm": 0.12260302156209946, "learning_rate": 2.3902392286062764e-05, "loss": 0.0256, "step": 110570 }, { "epoch": 2.209743815195236, "grad_norm": 0.11357271671295166, "learning_rate": 2.3893578043370707e-05, "loss": 0.0251, "step": 110580 }, { "epoch": 2.2099436473362375, "grad_norm": 0.08703719824552536, "learning_rate": 2.3884764915897028e-05, "loss": 0.0213, "step": 110590 }, { "epoch": 2.210143479477239, "grad_norm": 0.16085335612297058, "learning_rate": 2.3875952904018222e-05, "loss": 0.0302, "step": 110600 }, { "epoch": 2.2103433116182405, "grad_norm": 0.12515278160572052, "learning_rate": 2.3867142008110722e-05, "loss": 0.0272, "step": 110610 }, { "epoch": 2.210543143759242, "grad_norm": 0.11169399321079254, "learning_rate": 2.385833222855092e-05, "loss": 0.0258, "step": 110620 }, { "epoch": 2.210742975900244, "grad_norm": 0.12381517142057419, "learning_rate": 2.3849523565715127e-05, "loss": 0.034, "step": 110630 }, { "epoch": 2.210942808041245, "grad_norm": 0.12488140910863876, "learning_rate": 2.384071601997962e-05, "loss": 0.0274, "step": 110640 }, { "epoch": 2.211142640182247, "grad_norm": 0.1724618375301361, "learning_rate": 2.3831909591720713e-05, "loss": 0.023, "step": 110650 }, { "epoch": 2.2113424723232487, "grad_norm": 0.10598427057266235, "learning_rate": 2.3823104281314536e-05, "loss": 0.0258, "step": 110660 }, { "epoch": 2.21154230446425, "grad_norm": 0.11898968368768692, "learning_rate": 2.3814300089137258e-05, "loss": 0.0243, "step": 110670 }, { "epoch": 2.2117421366052517, "grad_norm": 0.1837163120508194, "learning_rate": 2.3805497015564972e-05, "loss": 0.0297, "step": 110680 }, { "epoch": 2.2119419687462534, "grad_norm": 0.11207439750432968, "learning_rate": 2.3796695060973733e-05, "loss": 0.0252, "step": 110690 }, { "epoch": 2.2121418008872547, "grad_norm": 0.1358693242073059, "learning_rate": 2.3787894225739542e-05, "loss": 0.0266, "step": 110700 }, { "epoch": 2.2123416330282564, "grad_norm": 0.09659167379140854, "learning_rate": 2.3779094510238357e-05, "loss": 0.0261, "step": 110710 }, { "epoch": 2.2125414651692576, "grad_norm": 0.15506407618522644, "learning_rate": 2.3770295914846115e-05, "loss": 0.0264, "step": 110720 }, { "epoch": 2.2127412973102594, "grad_norm": 0.15348351001739502, "learning_rate": 2.3761498439938597e-05, "loss": 0.0348, "step": 110730 }, { "epoch": 2.212941129451261, "grad_norm": 0.14889180660247803, "learning_rate": 2.375270208589168e-05, "loss": 0.0255, "step": 110740 }, { "epoch": 2.2131409615922624, "grad_norm": 0.13572216033935547, "learning_rate": 2.374390685308114e-05, "loss": 0.0249, "step": 110750 }, { "epoch": 2.213340793733264, "grad_norm": 0.11609984189271927, "learning_rate": 2.3735112741882642e-05, "loss": 0.023, "step": 110760 }, { "epoch": 2.213540625874266, "grad_norm": 0.11869470030069351, "learning_rate": 2.372631975267186e-05, "loss": 0.0261, "step": 110770 }, { "epoch": 2.213740458015267, "grad_norm": 0.10207698494195938, "learning_rate": 2.371752788582447e-05, "loss": 0.026, "step": 110780 }, { "epoch": 2.213940290156269, "grad_norm": 0.11119247227907181, "learning_rate": 2.3708737141715988e-05, "loss": 0.0217, "step": 110790 }, { "epoch": 2.21414012229727, "grad_norm": 0.11570774018764496, "learning_rate": 2.369994752072195e-05, "loss": 0.0229, "step": 110800 }, { "epoch": 2.214339954438272, "grad_norm": 0.21288533508777618, "learning_rate": 2.3691159023217846e-05, "loss": 0.0274, "step": 110810 }, { "epoch": 2.2145397865792735, "grad_norm": 0.14164163172245026, "learning_rate": 2.3682371649579094e-05, "loss": 0.0248, "step": 110820 }, { "epoch": 2.214739618720275, "grad_norm": 0.09935145825147629, "learning_rate": 2.367358540018108e-05, "loss": 0.0277, "step": 110830 }, { "epoch": 2.2149394508612765, "grad_norm": 0.11769824475049973, "learning_rate": 2.3664800275399137e-05, "loss": 0.0312, "step": 110840 }, { "epoch": 2.2151392830022782, "grad_norm": 0.1052452102303505, "learning_rate": 2.3656016275608545e-05, "loss": 0.0306, "step": 110850 }, { "epoch": 2.2153391151432795, "grad_norm": 0.21124783158302307, "learning_rate": 2.3647233401184538e-05, "loss": 0.0311, "step": 110860 }, { "epoch": 2.2155389472842812, "grad_norm": 0.14485812187194824, "learning_rate": 2.3638451652502314e-05, "loss": 0.0278, "step": 110870 }, { "epoch": 2.215738779425283, "grad_norm": 0.10966459661722183, "learning_rate": 2.3629671029937006e-05, "loss": 0.0366, "step": 110880 }, { "epoch": 2.2159386115662842, "grad_norm": 0.11012914031744003, "learning_rate": 2.3620891533863726e-05, "loss": 0.0278, "step": 110890 }, { "epoch": 2.216138443707286, "grad_norm": 0.11172939836978912, "learning_rate": 2.361211316465748e-05, "loss": 0.0225, "step": 110900 }, { "epoch": 2.2163382758482872, "grad_norm": 0.11460082232952118, "learning_rate": 2.3603335922693265e-05, "loss": 0.0269, "step": 110910 }, { "epoch": 2.216538107989289, "grad_norm": 0.1013457328081131, "learning_rate": 2.359455980834609e-05, "loss": 0.0238, "step": 110920 }, { "epoch": 2.2167379401302907, "grad_norm": 0.160341277718544, "learning_rate": 2.3585784821990787e-05, "loss": 0.0283, "step": 110930 }, { "epoch": 2.216937772271292, "grad_norm": 0.12747563421726227, "learning_rate": 2.3577010964002233e-05, "loss": 0.0234, "step": 110940 }, { "epoch": 2.2171376044122937, "grad_norm": 0.15616007149219513, "learning_rate": 2.356823823475523e-05, "loss": 0.0236, "step": 110950 }, { "epoch": 2.2173374365532954, "grad_norm": 0.12204093486070633, "learning_rate": 2.3559466634624532e-05, "loss": 0.0292, "step": 110960 }, { "epoch": 2.2175372686942967, "grad_norm": 0.09829725325107574, "learning_rate": 2.3550696163984847e-05, "loss": 0.0208, "step": 110970 }, { "epoch": 2.2177371008352984, "grad_norm": 0.10735278576612473, "learning_rate": 2.3541926823210837e-05, "loss": 0.0325, "step": 110980 }, { "epoch": 2.2179369329762997, "grad_norm": 0.14331497251987457, "learning_rate": 2.3533158612677124e-05, "loss": 0.0257, "step": 110990 }, { "epoch": 2.2181367651173014, "grad_norm": 0.14949390292167664, "learning_rate": 2.3524391532758218e-05, "loss": 0.0291, "step": 111000 }, { "epoch": 2.218336597258303, "grad_norm": 0.12909157574176788, "learning_rate": 2.351562558382868e-05, "loss": 0.0247, "step": 111010 }, { "epoch": 2.2185364293993044, "grad_norm": 0.12547054886817932, "learning_rate": 2.350686076626299e-05, "loss": 0.027, "step": 111020 }, { "epoch": 2.218736261540306, "grad_norm": 0.149956613779068, "learning_rate": 2.3498097080435522e-05, "loss": 0.0315, "step": 111030 }, { "epoch": 2.218936093681308, "grad_norm": 0.09027580171823502, "learning_rate": 2.3489334526720637e-05, "loss": 0.0283, "step": 111040 }, { "epoch": 2.219135925822309, "grad_norm": 0.12437359988689423, "learning_rate": 2.3480573105492715e-05, "loss": 0.0214, "step": 111050 }, { "epoch": 2.219335757963311, "grad_norm": 0.08229803293943405, "learning_rate": 2.3471812817125976e-05, "loss": 0.0202, "step": 111060 }, { "epoch": 2.2195355901043126, "grad_norm": 0.08810123801231384, "learning_rate": 2.346305366199466e-05, "loss": 0.0304, "step": 111070 }, { "epoch": 2.219735422245314, "grad_norm": 0.1098538190126419, "learning_rate": 2.3454295640472945e-05, "loss": 0.0176, "step": 111080 }, { "epoch": 2.2199352543863156, "grad_norm": 0.13790404796600342, "learning_rate": 2.3445538752934948e-05, "loss": 0.0267, "step": 111090 }, { "epoch": 2.220135086527317, "grad_norm": 0.11139920353889465, "learning_rate": 2.343678299975476e-05, "loss": 0.0324, "step": 111100 }, { "epoch": 2.2203349186683186, "grad_norm": 0.1224493756890297, "learning_rate": 2.34280283813064e-05, "loss": 0.0265, "step": 111110 }, { "epoch": 2.2205347508093203, "grad_norm": 0.15520116686820984, "learning_rate": 2.3419274897963873e-05, "loss": 0.0233, "step": 111120 }, { "epoch": 2.2207345829503216, "grad_norm": 0.1404053121805191, "learning_rate": 2.341052255010105e-05, "loss": 0.032, "step": 111130 }, { "epoch": 2.2209344150913233, "grad_norm": 0.11218860745429993, "learning_rate": 2.3401771338091884e-05, "loss": 0.0273, "step": 111140 }, { "epoch": 2.221134247232325, "grad_norm": 0.15831123292446136, "learning_rate": 2.3393021262310195e-05, "loss": 0.0261, "step": 111150 }, { "epoch": 2.2213340793733263, "grad_norm": 0.11376405507326126, "learning_rate": 2.3384272323129746e-05, "loss": 0.0264, "step": 111160 }, { "epoch": 2.221533911514328, "grad_norm": 0.14520838856697083, "learning_rate": 2.3375524520924286e-05, "loss": 0.041, "step": 111170 }, { "epoch": 2.2217337436553297, "grad_norm": 0.10677237808704376, "learning_rate": 2.336677785606749e-05, "loss": 0.0242, "step": 111180 }, { "epoch": 2.221933575796331, "grad_norm": 0.10589136928319931, "learning_rate": 2.3358032328933054e-05, "loss": 0.0263, "step": 111190 }, { "epoch": 2.2221334079373327, "grad_norm": 0.12776978313922882, "learning_rate": 2.3349287939894516e-05, "loss": 0.0263, "step": 111200 }, { "epoch": 2.222333240078334, "grad_norm": 0.10308778285980225, "learning_rate": 2.334054468932544e-05, "loss": 0.0258, "step": 111210 }, { "epoch": 2.2225330722193357, "grad_norm": 0.13403552770614624, "learning_rate": 2.3331802577599317e-05, "loss": 0.0205, "step": 111220 }, { "epoch": 2.2227329043603374, "grad_norm": 0.1103883683681488, "learning_rate": 2.3323061605089597e-05, "loss": 0.0281, "step": 111230 }, { "epoch": 2.2229327365013387, "grad_norm": 0.09233991801738739, "learning_rate": 2.3314321772169677e-05, "loss": 0.03, "step": 111240 }, { "epoch": 2.2231325686423404, "grad_norm": 0.12576039135456085, "learning_rate": 2.330558307921291e-05, "loss": 0.029, "step": 111250 }, { "epoch": 2.223332400783342, "grad_norm": 0.10396023094654083, "learning_rate": 2.329684552659259e-05, "loss": 0.0256, "step": 111260 }, { "epoch": 2.2235322329243434, "grad_norm": 0.13514919579029083, "learning_rate": 2.3288109114681982e-05, "loss": 0.0373, "step": 111270 }, { "epoch": 2.223732065065345, "grad_norm": 0.13311368227005005, "learning_rate": 2.327937384385428e-05, "loss": 0.0252, "step": 111280 }, { "epoch": 2.223931897206347, "grad_norm": 0.14797961711883545, "learning_rate": 2.3270639714482656e-05, "loss": 0.0312, "step": 111290 }, { "epoch": 2.224131729347348, "grad_norm": 0.14965495467185974, "learning_rate": 2.326190672694018e-05, "loss": 0.0263, "step": 111300 }, { "epoch": 2.22433156148835, "grad_norm": 0.12662667036056519, "learning_rate": 2.325317488159991e-05, "loss": 0.0285, "step": 111310 }, { "epoch": 2.224531393629351, "grad_norm": 0.11665286868810654, "learning_rate": 2.324444417883491e-05, "loss": 0.0248, "step": 111320 }, { "epoch": 2.224731225770353, "grad_norm": 0.10710202157497406, "learning_rate": 2.323571461901809e-05, "loss": 0.0173, "step": 111330 }, { "epoch": 2.2249310579113546, "grad_norm": 0.1500457227230072, "learning_rate": 2.322698620252237e-05, "loss": 0.0205, "step": 111340 }, { "epoch": 2.225130890052356, "grad_norm": 0.17716370522975922, "learning_rate": 2.321825892972062e-05, "loss": 0.0195, "step": 111350 }, { "epoch": 2.2253307221933576, "grad_norm": 0.10145372897386551, "learning_rate": 2.3209532800985644e-05, "loss": 0.0272, "step": 111360 }, { "epoch": 2.2255305543343593, "grad_norm": 0.11347056925296783, "learning_rate": 2.3200807816690214e-05, "loss": 0.0226, "step": 111370 }, { "epoch": 2.2257303864753606, "grad_norm": 0.1327623724937439, "learning_rate": 2.3192083977207042e-05, "loss": 0.0306, "step": 111380 }, { "epoch": 2.2259302186163623, "grad_norm": 0.12337809056043625, "learning_rate": 2.318336128290881e-05, "loss": 0.0258, "step": 111390 }, { "epoch": 2.226130050757364, "grad_norm": 0.10937874764204025, "learning_rate": 2.3174639734168084e-05, "loss": 0.0196, "step": 111400 }, { "epoch": 2.2263298828983653, "grad_norm": 0.11919985711574554, "learning_rate": 2.3165919331357493e-05, "loss": 0.0257, "step": 111410 }, { "epoch": 2.226529715039367, "grad_norm": 0.09465046972036362, "learning_rate": 2.3157200074849543e-05, "loss": 0.0183, "step": 111420 }, { "epoch": 2.2267295471803683, "grad_norm": 0.20395393669605255, "learning_rate": 2.3148481965016683e-05, "loss": 0.0264, "step": 111430 }, { "epoch": 2.22692937932137, "grad_norm": 0.13349087536334991, "learning_rate": 2.313976500223132e-05, "loss": 0.026, "step": 111440 }, { "epoch": 2.2271292114623717, "grad_norm": 0.14604052901268005, "learning_rate": 2.31310491868659e-05, "loss": 0.0266, "step": 111450 }, { "epoch": 2.227329043603373, "grad_norm": 0.14883551001548767, "learning_rate": 2.312233451929267e-05, "loss": 0.0241, "step": 111460 }, { "epoch": 2.2275288757443747, "grad_norm": 0.11038954555988312, "learning_rate": 2.311362099988394e-05, "loss": 0.0236, "step": 111470 }, { "epoch": 2.2277287078853765, "grad_norm": 0.10355237871408463, "learning_rate": 2.3104908629011923e-05, "loss": 0.0247, "step": 111480 }, { "epoch": 2.2279285400263777, "grad_norm": 0.09628233313560486, "learning_rate": 2.3096197407048804e-05, "loss": 0.0302, "step": 111490 }, { "epoch": 2.2281283721673795, "grad_norm": 0.11846980452537537, "learning_rate": 2.308748733436671e-05, "loss": 0.0284, "step": 111500 }, { "epoch": 2.228328204308381, "grad_norm": 0.12196565419435501, "learning_rate": 2.3078778411337715e-05, "loss": 0.0282, "step": 111510 }, { "epoch": 2.2285280364493825, "grad_norm": 0.12974737584590912, "learning_rate": 2.3070070638333867e-05, "loss": 0.0268, "step": 111520 }, { "epoch": 2.228727868590384, "grad_norm": 0.12258991599082947, "learning_rate": 2.3061364015727095e-05, "loss": 0.022, "step": 111530 }, { "epoch": 2.2289277007313855, "grad_norm": 0.13283108174800873, "learning_rate": 2.305265854388938e-05, "loss": 0.0259, "step": 111540 }, { "epoch": 2.229127532872387, "grad_norm": 0.14603053033351898, "learning_rate": 2.3043954223192615e-05, "loss": 0.0294, "step": 111550 }, { "epoch": 2.229327365013389, "grad_norm": 0.1486007124185562, "learning_rate": 2.3035251054008578e-05, "loss": 0.0309, "step": 111560 }, { "epoch": 2.22952719715439, "grad_norm": 0.13250933587551117, "learning_rate": 2.3026549036709088e-05, "loss": 0.0359, "step": 111570 }, { "epoch": 2.229727029295392, "grad_norm": 0.09733381867408752, "learning_rate": 2.3017848171665846e-05, "loss": 0.0294, "step": 111580 }, { "epoch": 2.2299268614363936, "grad_norm": 0.1253996193408966, "learning_rate": 2.3009148459250603e-05, "loss": 0.0267, "step": 111590 }, { "epoch": 2.230126693577395, "grad_norm": 0.12530885636806488, "learning_rate": 2.3000449899834936e-05, "loss": 0.0248, "step": 111600 }, { "epoch": 2.2303265257183966, "grad_norm": 0.1656452715396881, "learning_rate": 2.2991752493790448e-05, "loss": 0.0272, "step": 111610 }, { "epoch": 2.230526357859398, "grad_norm": 0.16609084606170654, "learning_rate": 2.2983056241488676e-05, "loss": 0.0267, "step": 111620 }, { "epoch": 2.2307261900003996, "grad_norm": 0.14378070831298828, "learning_rate": 2.297436114330111e-05, "loss": 0.0319, "step": 111630 }, { "epoch": 2.2309260221414013, "grad_norm": 0.12175451219081879, "learning_rate": 2.296566719959919e-05, "loss": 0.0297, "step": 111640 }, { "epoch": 2.2311258542824026, "grad_norm": 0.11476706713438034, "learning_rate": 2.2956974410754306e-05, "loss": 0.0376, "step": 111650 }, { "epoch": 2.2313256864234043, "grad_norm": 0.1710778921842575, "learning_rate": 2.2948282777137807e-05, "loss": 0.021, "step": 111660 }, { "epoch": 2.231525518564406, "grad_norm": 0.13506972789764404, "learning_rate": 2.2939592299120942e-05, "loss": 0.0269, "step": 111670 }, { "epoch": 2.2317253507054073, "grad_norm": 0.11407482624053955, "learning_rate": 2.2930902977074998e-05, "loss": 0.0366, "step": 111680 }, { "epoch": 2.231925182846409, "grad_norm": 0.09979791939258575, "learning_rate": 2.2922214811371173e-05, "loss": 0.0221, "step": 111690 }, { "epoch": 2.232125014987411, "grad_norm": 0.12141553312540054, "learning_rate": 2.2913527802380563e-05, "loss": 0.0214, "step": 111700 }, { "epoch": 2.232324847128412, "grad_norm": 0.10965949296951294, "learning_rate": 2.290484195047427e-05, "loss": 0.0224, "step": 111710 }, { "epoch": 2.232524679269414, "grad_norm": 0.10576558858156204, "learning_rate": 2.2896157256023394e-05, "loss": 0.0264, "step": 111720 }, { "epoch": 2.232724511410415, "grad_norm": 0.12990690767765045, "learning_rate": 2.2887473719398865e-05, "loss": 0.0256, "step": 111730 }, { "epoch": 2.232924343551417, "grad_norm": 0.11612562090158463, "learning_rate": 2.2878791340971657e-05, "loss": 0.0205, "step": 111740 }, { "epoch": 2.2331241756924185, "grad_norm": 0.0774167850613594, "learning_rate": 2.2870110121112654e-05, "loss": 0.0267, "step": 111750 }, { "epoch": 2.2333240078334198, "grad_norm": 0.13333579897880554, "learning_rate": 2.2861430060192712e-05, "loss": 0.0272, "step": 111760 }, { "epoch": 2.2335238399744215, "grad_norm": 0.13705189526081085, "learning_rate": 2.2852751158582618e-05, "loss": 0.0325, "step": 111770 }, { "epoch": 2.233723672115423, "grad_norm": 0.08680310100317001, "learning_rate": 2.2844073416653128e-05, "loss": 0.0217, "step": 111780 }, { "epoch": 2.2339235042564245, "grad_norm": 0.10763157904148102, "learning_rate": 2.283539683477495e-05, "loss": 0.0247, "step": 111790 }, { "epoch": 2.234123336397426, "grad_norm": 0.1334955394268036, "learning_rate": 2.2826721413318682e-05, "loss": 0.0267, "step": 111800 }, { "epoch": 2.2343231685384275, "grad_norm": 0.11066009104251862, "learning_rate": 2.281804715265497e-05, "loss": 0.0283, "step": 111810 }, { "epoch": 2.234523000679429, "grad_norm": 0.1195305660367012, "learning_rate": 2.2809374053154375e-05, "loss": 0.0269, "step": 111820 }, { "epoch": 2.234722832820431, "grad_norm": 0.18269477784633636, "learning_rate": 2.2800702115187345e-05, "loss": 0.0281, "step": 111830 }, { "epoch": 2.234922664961432, "grad_norm": 0.1268240362405777, "learning_rate": 2.2792031339124355e-05, "loss": 0.0327, "step": 111840 }, { "epoch": 2.235122497102434, "grad_norm": 0.14688953757286072, "learning_rate": 2.2783361725335806e-05, "loss": 0.0234, "step": 111850 }, { "epoch": 2.2353223292434357, "grad_norm": 0.08887235075235367, "learning_rate": 2.2774693274192044e-05, "loss": 0.0249, "step": 111860 }, { "epoch": 2.235522161384437, "grad_norm": 0.131890207529068, "learning_rate": 2.2766025986063367e-05, "loss": 0.0222, "step": 111870 }, { "epoch": 2.2357219935254387, "grad_norm": 0.12454258650541306, "learning_rate": 2.2757359861320033e-05, "loss": 0.0255, "step": 111880 }, { "epoch": 2.2359218256664404, "grad_norm": 0.16626766324043274, "learning_rate": 2.274869490033224e-05, "loss": 0.0248, "step": 111890 }, { "epoch": 2.2361216578074417, "grad_norm": 0.1647413671016693, "learning_rate": 2.2740031103470137e-05, "loss": 0.0311, "step": 111900 }, { "epoch": 2.2363214899484434, "grad_norm": 0.08195873349905014, "learning_rate": 2.273136847110383e-05, "loss": 0.0224, "step": 111910 }, { "epoch": 2.2365213220894447, "grad_norm": 0.1082511618733406, "learning_rate": 2.2722707003603384e-05, "loss": 0.0177, "step": 111920 }, { "epoch": 2.2367211542304464, "grad_norm": 0.1573808193206787, "learning_rate": 2.271404670133877e-05, "loss": 0.0268, "step": 111930 }, { "epoch": 2.236920986371448, "grad_norm": 0.12078101187944412, "learning_rate": 2.270538756467993e-05, "loss": 0.0258, "step": 111940 }, { "epoch": 2.2371208185124494, "grad_norm": 0.11263900250196457, "learning_rate": 2.2696729593996835e-05, "loss": 0.0279, "step": 111950 }, { "epoch": 2.237320650653451, "grad_norm": 0.08827149122953415, "learning_rate": 2.268807278965927e-05, "loss": 0.0332, "step": 111960 }, { "epoch": 2.237520482794453, "grad_norm": 0.10661017894744873, "learning_rate": 2.2679417152037062e-05, "loss": 0.0295, "step": 111970 }, { "epoch": 2.237720314935454, "grad_norm": 0.11586993932723999, "learning_rate": 2.267076268149995e-05, "loss": 0.0288, "step": 111980 }, { "epoch": 2.237920147076456, "grad_norm": 0.13867539167404175, "learning_rate": 2.2662109378417684e-05, "loss": 0.0332, "step": 111990 }, { "epoch": 2.2381199792174575, "grad_norm": 0.09318647533655167, "learning_rate": 2.2653457243159864e-05, "loss": 0.0219, "step": 112000 }, { "epoch": 2.238319811358459, "grad_norm": 0.13795268535614014, "learning_rate": 2.264480627609612e-05, "loss": 0.0314, "step": 112010 }, { "epoch": 2.2385196434994605, "grad_norm": 0.12654639780521393, "learning_rate": 2.2636156477595995e-05, "loss": 0.0268, "step": 112020 }, { "epoch": 2.238719475640462, "grad_norm": 0.1137855052947998, "learning_rate": 2.2627507848028996e-05, "loss": 0.0309, "step": 112030 }, { "epoch": 2.2389193077814635, "grad_norm": 0.14206968247890472, "learning_rate": 2.261886038776458e-05, "loss": 0.0221, "step": 112040 }, { "epoch": 2.2391191399224653, "grad_norm": 0.11804486066102982, "learning_rate": 2.261021409717214e-05, "loss": 0.0269, "step": 112050 }, { "epoch": 2.2393189720634665, "grad_norm": 0.10666366666555405, "learning_rate": 2.2601568976621063e-05, "loss": 0.0226, "step": 112060 }, { "epoch": 2.2395188042044682, "grad_norm": 0.11676064133644104, "learning_rate": 2.259292502648058e-05, "loss": 0.0292, "step": 112070 }, { "epoch": 2.23971863634547, "grad_norm": 0.1491880714893341, "learning_rate": 2.258428224712002e-05, "loss": 0.0227, "step": 112080 }, { "epoch": 2.2399184684864712, "grad_norm": 0.12005545198917389, "learning_rate": 2.2575640638908574e-05, "loss": 0.0257, "step": 112090 }, { "epoch": 2.240118300627473, "grad_norm": 0.1575389802455902, "learning_rate": 2.256700020221536e-05, "loss": 0.0221, "step": 112100 }, { "epoch": 2.2403181327684747, "grad_norm": 0.10616962611675262, "learning_rate": 2.2558360937409507e-05, "loss": 0.019, "step": 112110 }, { "epoch": 2.240517964909476, "grad_norm": 0.10368672013282776, "learning_rate": 2.2549722844860056e-05, "loss": 0.0236, "step": 112120 }, { "epoch": 2.2407177970504777, "grad_norm": 0.11451409757137299, "learning_rate": 2.2541085924936017e-05, "loss": 0.0288, "step": 112130 }, { "epoch": 2.240917629191479, "grad_norm": 0.1298474818468094, "learning_rate": 2.2532450178006347e-05, "loss": 0.0296, "step": 112140 }, { "epoch": 2.2411174613324807, "grad_norm": 0.1136862188577652, "learning_rate": 2.2523815604439947e-05, "loss": 0.0266, "step": 112150 }, { "epoch": 2.2413172934734824, "grad_norm": 0.10444293916225433, "learning_rate": 2.2515182204605662e-05, "loss": 0.0308, "step": 112160 }, { "epoch": 2.2415171256144837, "grad_norm": 0.1329329013824463, "learning_rate": 2.250654997887231e-05, "loss": 0.0253, "step": 112170 }, { "epoch": 2.2417169577554854, "grad_norm": 0.12073225528001785, "learning_rate": 2.249791892760863e-05, "loss": 0.0202, "step": 112180 }, { "epoch": 2.241916789896487, "grad_norm": 0.11264370381832123, "learning_rate": 2.2489289051183353e-05, "loss": 0.0336, "step": 112190 }, { "epoch": 2.2421166220374884, "grad_norm": 0.08284693211317062, "learning_rate": 2.248066034996509e-05, "loss": 0.0291, "step": 112200 }, { "epoch": 2.24231645417849, "grad_norm": 0.10931002348661423, "learning_rate": 2.247203282432244e-05, "loss": 0.0333, "step": 112210 }, { "epoch": 2.242516286319492, "grad_norm": 0.14137345552444458, "learning_rate": 2.246340647462402e-05, "loss": 0.0297, "step": 112220 }, { "epoch": 2.242716118460493, "grad_norm": 0.11327441781759262, "learning_rate": 2.2454781301238264e-05, "loss": 0.0227, "step": 112230 }, { "epoch": 2.242915950601495, "grad_norm": 0.1428440362215042, "learning_rate": 2.2446157304533655e-05, "loss": 0.0284, "step": 112240 }, { "epoch": 2.243115782742496, "grad_norm": 0.1133628636598587, "learning_rate": 2.2437534484878586e-05, "loss": 0.0159, "step": 112250 }, { "epoch": 2.243315614883498, "grad_norm": 0.09435072541236877, "learning_rate": 2.2428912842641408e-05, "loss": 0.0314, "step": 112260 }, { "epoch": 2.2435154470244996, "grad_norm": 0.11359430849552155, "learning_rate": 2.242029237819042e-05, "loss": 0.0237, "step": 112270 }, { "epoch": 2.243715279165501, "grad_norm": 0.1629675030708313, "learning_rate": 2.2411673091893887e-05, "loss": 0.0258, "step": 112280 }, { "epoch": 2.2439151113065026, "grad_norm": 0.10207336395978928, "learning_rate": 2.240305498412001e-05, "loss": 0.0231, "step": 112290 }, { "epoch": 2.2441149434475043, "grad_norm": 0.10555162280797958, "learning_rate": 2.239443805523689e-05, "loss": 0.0306, "step": 112300 }, { "epoch": 2.2443147755885056, "grad_norm": 0.11544192582368851, "learning_rate": 2.238582230561268e-05, "loss": 0.0304, "step": 112310 }, { "epoch": 2.2445146077295073, "grad_norm": 0.11769559979438782, "learning_rate": 2.2377207735615435e-05, "loss": 0.0227, "step": 112320 }, { "epoch": 2.244714439870509, "grad_norm": 0.1686064749956131, "learning_rate": 2.236859434561311e-05, "loss": 0.0227, "step": 112330 }, { "epoch": 2.2449142720115103, "grad_norm": 0.10444402694702148, "learning_rate": 2.2359982135973657e-05, "loss": 0.0255, "step": 112340 }, { "epoch": 2.245114104152512, "grad_norm": 0.14844153821468353, "learning_rate": 2.2351371107065024e-05, "loss": 0.0294, "step": 112350 }, { "epoch": 2.2453139362935133, "grad_norm": 0.15232987701892853, "learning_rate": 2.234276125925501e-05, "loss": 0.0393, "step": 112360 }, { "epoch": 2.245513768434515, "grad_norm": 0.14145222306251526, "learning_rate": 2.2334152592911424e-05, "loss": 0.0341, "step": 112370 }, { "epoch": 2.2457136005755167, "grad_norm": 0.11984491348266602, "learning_rate": 2.2325545108402018e-05, "loss": 0.0215, "step": 112380 }, { "epoch": 2.245913432716518, "grad_norm": 0.12209921330213547, "learning_rate": 2.2316938806094483e-05, "loss": 0.0233, "step": 112390 }, { "epoch": 2.2461132648575197, "grad_norm": 0.12883563339710236, "learning_rate": 2.230833368635647e-05, "loss": 0.0211, "step": 112400 }, { "epoch": 2.2463130969985214, "grad_norm": 0.13949759304523468, "learning_rate": 2.229972974955557e-05, "loss": 0.041, "step": 112410 }, { "epoch": 2.2465129291395227, "grad_norm": 0.10260568559169769, "learning_rate": 2.2291126996059326e-05, "loss": 0.0227, "step": 112420 }, { "epoch": 2.2467127612805244, "grad_norm": 0.10075412690639496, "learning_rate": 2.2282525426235233e-05, "loss": 0.0298, "step": 112430 }, { "epoch": 2.2469125934215257, "grad_norm": 0.10347907990217209, "learning_rate": 2.2273925040450745e-05, "loss": 0.0256, "step": 112440 }, { "epoch": 2.2471124255625274, "grad_norm": 0.10141617804765701, "learning_rate": 2.226532583907323e-05, "loss": 0.0287, "step": 112450 }, { "epoch": 2.247312257703529, "grad_norm": 0.12943528592586517, "learning_rate": 2.2256727822470076e-05, "loss": 0.0226, "step": 112460 }, { "epoch": 2.2475120898445304, "grad_norm": 0.1419111043214798, "learning_rate": 2.2248130991008497e-05, "loss": 0.027, "step": 112470 }, { "epoch": 2.247711921985532, "grad_norm": 0.12617751955986023, "learning_rate": 2.2239535345055807e-05, "loss": 0.0281, "step": 112480 }, { "epoch": 2.247911754126534, "grad_norm": 0.1272231936454773, "learning_rate": 2.2230940884979183e-05, "loss": 0.0196, "step": 112490 }, { "epoch": 2.248111586267535, "grad_norm": 0.14310921728610992, "learning_rate": 2.2222347611145734e-05, "loss": 0.0349, "step": 112500 }, { "epoch": 2.248311418408537, "grad_norm": 0.16573214530944824, "learning_rate": 2.2213755523922564e-05, "loss": 0.0303, "step": 112510 }, { "epoch": 2.248511250549538, "grad_norm": 0.2389514446258545, "learning_rate": 2.2205164623676715e-05, "loss": 0.0257, "step": 112520 }, { "epoch": 2.24871108269054, "grad_norm": 0.09650103002786636, "learning_rate": 2.219657491077517e-05, "loss": 0.0263, "step": 112530 }, { "epoch": 2.2489109148315416, "grad_norm": 0.10798971354961395, "learning_rate": 2.2187986385584868e-05, "loss": 0.0363, "step": 112540 }, { "epoch": 2.249110746972543, "grad_norm": 0.14607185125350952, "learning_rate": 2.2179399048472697e-05, "loss": 0.0243, "step": 112550 }, { "epoch": 2.2493105791135446, "grad_norm": 0.1398203819990158, "learning_rate": 2.2170812899805493e-05, "loss": 0.0236, "step": 112560 }, { "epoch": 2.2495104112545463, "grad_norm": 0.11821113526821136, "learning_rate": 2.2162227939950036e-05, "loss": 0.0204, "step": 112570 }, { "epoch": 2.2497102433955476, "grad_norm": 0.1270511895418167, "learning_rate": 2.2153644169273064e-05, "loss": 0.0241, "step": 112580 }, { "epoch": 2.2499100755365493, "grad_norm": 0.0994282066822052, "learning_rate": 2.2145061588141273e-05, "loss": 0.0276, "step": 112590 }, { "epoch": 2.250109907677551, "grad_norm": 0.0832107812166214, "learning_rate": 2.2136480196921268e-05, "loss": 0.033, "step": 112600 }, { "epoch": 2.2503097398185523, "grad_norm": 0.12555436789989471, "learning_rate": 2.212789999597961e-05, "loss": 0.0262, "step": 112610 }, { "epoch": 2.250509571959554, "grad_norm": 0.1102781817317009, "learning_rate": 2.211932098568291e-05, "loss": 0.023, "step": 112620 }, { "epoch": 2.2507094041005553, "grad_norm": 0.14624086022377014, "learning_rate": 2.211074316639758e-05, "loss": 0.0294, "step": 112630 }, { "epoch": 2.250909236241557, "grad_norm": 0.11647181957960129, "learning_rate": 2.210216653849006e-05, "loss": 0.0266, "step": 112640 }, { "epoch": 2.2511090683825588, "grad_norm": 0.1230878084897995, "learning_rate": 2.2093591102326744e-05, "loss": 0.0264, "step": 112650 }, { "epoch": 2.25130890052356, "grad_norm": 0.12137014418840408, "learning_rate": 2.2085016858273945e-05, "loss": 0.0294, "step": 112660 }, { "epoch": 2.2515087326645618, "grad_norm": 0.11810802668333054, "learning_rate": 2.2076443806697945e-05, "loss": 0.0295, "step": 112670 }, { "epoch": 2.2517085648055635, "grad_norm": 0.09467728435993195, "learning_rate": 2.206787194796497e-05, "loss": 0.021, "step": 112680 }, { "epoch": 2.2519083969465647, "grad_norm": 0.12493815273046494, "learning_rate": 2.205930128244121e-05, "loss": 0.0318, "step": 112690 }, { "epoch": 2.2521082290875665, "grad_norm": 0.11053944379091263, "learning_rate": 2.2050731810492736e-05, "loss": 0.0285, "step": 112700 }, { "epoch": 2.252308061228568, "grad_norm": 0.12791861593723297, "learning_rate": 2.2042163532485673e-05, "loss": 0.0268, "step": 112710 }, { "epoch": 2.2525078933695695, "grad_norm": 0.1440083533525467, "learning_rate": 2.2033596448786036e-05, "loss": 0.021, "step": 112720 }, { "epoch": 2.252707725510571, "grad_norm": 0.11531740427017212, "learning_rate": 2.2025030559759767e-05, "loss": 0.024, "step": 112730 }, { "epoch": 2.2529075576515725, "grad_norm": 0.15953603386878967, "learning_rate": 2.2016465865772785e-05, "loss": 0.0266, "step": 112740 }, { "epoch": 2.253107389792574, "grad_norm": 0.13635094463825226, "learning_rate": 2.2007902367191007e-05, "loss": 0.0252, "step": 112750 }, { "epoch": 2.253307221933576, "grad_norm": 0.09272445738315582, "learning_rate": 2.1999340064380204e-05, "loss": 0.0276, "step": 112760 }, { "epoch": 2.253507054074577, "grad_norm": 0.1229863315820694, "learning_rate": 2.1990778957706153e-05, "loss": 0.0254, "step": 112770 }, { "epoch": 2.253706886215579, "grad_norm": 0.10907159000635147, "learning_rate": 2.198221904753457e-05, "loss": 0.0188, "step": 112780 }, { "epoch": 2.2539067183565806, "grad_norm": 0.08720149844884872, "learning_rate": 2.1973660334231117e-05, "loss": 0.0182, "step": 112790 }, { "epoch": 2.254106550497582, "grad_norm": 0.10716737061738968, "learning_rate": 2.196510281816141e-05, "loss": 0.02, "step": 112800 }, { "epoch": 2.2543063826385836, "grad_norm": 0.09681883454322815, "learning_rate": 2.1956546499691005e-05, "loss": 0.0195, "step": 112810 }, { "epoch": 2.2545062147795853, "grad_norm": 0.12183987349271774, "learning_rate": 2.1947991379185418e-05, "loss": 0.0297, "step": 112820 }, { "epoch": 2.2547060469205866, "grad_norm": 0.10338160395622253, "learning_rate": 2.1939437457010108e-05, "loss": 0.0301, "step": 112830 }, { "epoch": 2.2549058790615883, "grad_norm": 0.1362651288509369, "learning_rate": 2.193088473353048e-05, "loss": 0.0206, "step": 112840 }, { "epoch": 2.2551057112025896, "grad_norm": 0.13449348509311676, "learning_rate": 2.1922333209111895e-05, "loss": 0.0228, "step": 112850 }, { "epoch": 2.2553055433435913, "grad_norm": 0.11040971428155899, "learning_rate": 2.1913782884119666e-05, "loss": 0.0241, "step": 112860 }, { "epoch": 2.255505375484593, "grad_norm": 0.11433610320091248, "learning_rate": 2.1905233758919025e-05, "loss": 0.0213, "step": 112870 }, { "epoch": 2.2557052076255943, "grad_norm": 0.10175730288028717, "learning_rate": 2.1896685833875163e-05, "loss": 0.0255, "step": 112880 }, { "epoch": 2.255905039766596, "grad_norm": 0.11501416563987732, "learning_rate": 2.1888139109353297e-05, "loss": 0.0231, "step": 112890 }, { "epoch": 2.256104871907598, "grad_norm": 0.11640290915966034, "learning_rate": 2.1879593585718462e-05, "loss": 0.0229, "step": 112900 }, { "epoch": 2.256304704048599, "grad_norm": 0.13907332718372345, "learning_rate": 2.187104926333573e-05, "loss": 0.0257, "step": 112910 }, { "epoch": 2.256504536189601, "grad_norm": 0.13277025520801544, "learning_rate": 2.1862506142570098e-05, "loss": 0.0223, "step": 112920 }, { "epoch": 2.2567043683306025, "grad_norm": 0.13340915739536285, "learning_rate": 2.1853964223786517e-05, "loss": 0.0191, "step": 112930 }, { "epoch": 2.256904200471604, "grad_norm": 0.13152819871902466, "learning_rate": 2.1845423507349878e-05, "loss": 0.0247, "step": 112940 }, { "epoch": 2.2571040326126055, "grad_norm": 0.10171734541654587, "learning_rate": 2.1836883993625017e-05, "loss": 0.03, "step": 112950 }, { "epoch": 2.257303864753607, "grad_norm": 0.08663257211446762, "learning_rate": 2.1828345682976766e-05, "loss": 0.0229, "step": 112960 }, { "epoch": 2.2575036968946085, "grad_norm": 0.18203052878379822, "learning_rate": 2.181980857576979e-05, "loss": 0.0324, "step": 112970 }, { "epoch": 2.2577035290356102, "grad_norm": 0.11529635637998581, "learning_rate": 2.1811272672368844e-05, "loss": 0.0267, "step": 112980 }, { "epoch": 2.2579033611766115, "grad_norm": 0.14411558210849762, "learning_rate": 2.180273797313856e-05, "loss": 0.0224, "step": 112990 }, { "epoch": 2.258103193317613, "grad_norm": 0.1474786251783371, "learning_rate": 2.179420447844349e-05, "loss": 0.0309, "step": 113000 }, { "epoch": 2.258303025458615, "grad_norm": 0.11152289062738419, "learning_rate": 2.1785672188648177e-05, "loss": 0.0323, "step": 113010 }, { "epoch": 2.258502857599616, "grad_norm": 0.10038088262081146, "learning_rate": 2.177714110411715e-05, "loss": 0.0223, "step": 113020 }, { "epoch": 2.258702689740618, "grad_norm": 0.1522090882062912, "learning_rate": 2.1768611225214792e-05, "loss": 0.0357, "step": 113030 }, { "epoch": 2.2589025218816197, "grad_norm": 0.14911621809005737, "learning_rate": 2.1760082552305498e-05, "loss": 0.0348, "step": 113040 }, { "epoch": 2.259102354022621, "grad_norm": 0.12727496027946472, "learning_rate": 2.17515550857536e-05, "loss": 0.0314, "step": 113050 }, { "epoch": 2.2593021861636227, "grad_norm": 0.12251470983028412, "learning_rate": 2.1743028825923373e-05, "loss": 0.0266, "step": 113060 }, { "epoch": 2.259502018304624, "grad_norm": 0.10301247239112854, "learning_rate": 2.1734503773179045e-05, "loss": 0.0267, "step": 113070 }, { "epoch": 2.2597018504456257, "grad_norm": 0.10342273116111755, "learning_rate": 2.1725979927884794e-05, "loss": 0.0224, "step": 113080 }, { "epoch": 2.2599016825866274, "grad_norm": 0.1782691329717636, "learning_rate": 2.1717457290404754e-05, "loss": 0.0256, "step": 113090 }, { "epoch": 2.2601015147276287, "grad_norm": 0.08600908517837524, "learning_rate": 2.170893586110294e-05, "loss": 0.0205, "step": 113100 }, { "epoch": 2.2603013468686304, "grad_norm": 0.12494880706071854, "learning_rate": 2.170041564034344e-05, "loss": 0.0295, "step": 113110 }, { "epoch": 2.260501179009632, "grad_norm": 0.10702929645776749, "learning_rate": 2.1691896628490214e-05, "loss": 0.021, "step": 113120 }, { "epoch": 2.2607010111506334, "grad_norm": 0.12362607568502426, "learning_rate": 2.1683378825907142e-05, "loss": 0.028, "step": 113130 }, { "epoch": 2.260900843291635, "grad_norm": 0.10609735548496246, "learning_rate": 2.1674862232958104e-05, "loss": 0.0292, "step": 113140 }, { "epoch": 2.261100675432637, "grad_norm": 0.11194776743650436, "learning_rate": 2.166634685000692e-05, "loss": 0.0219, "step": 113150 }, { "epoch": 2.261300507573638, "grad_norm": 0.13290469348430634, "learning_rate": 2.1657832677417346e-05, "loss": 0.0259, "step": 113160 }, { "epoch": 2.26150033971464, "grad_norm": 0.18157045543193817, "learning_rate": 2.164931971555309e-05, "loss": 0.0223, "step": 113170 }, { "epoch": 2.261700171855641, "grad_norm": 0.09303642064332962, "learning_rate": 2.1640807964777822e-05, "loss": 0.0232, "step": 113180 }, { "epoch": 2.261900003996643, "grad_norm": 0.11181459575891495, "learning_rate": 2.1632297425455134e-05, "loss": 0.0258, "step": 113190 }, { "epoch": 2.2620998361376445, "grad_norm": 0.10629071295261383, "learning_rate": 2.1623788097948588e-05, "loss": 0.0238, "step": 113200 }, { "epoch": 2.262299668278646, "grad_norm": 0.0976899266242981, "learning_rate": 2.1615279982621684e-05, "loss": 0.0332, "step": 113210 }, { "epoch": 2.2624995004196475, "grad_norm": 0.11291079968214035, "learning_rate": 2.1606773079837877e-05, "loss": 0.0258, "step": 113220 }, { "epoch": 2.262699332560649, "grad_norm": 0.14478899538516998, "learning_rate": 2.1598267389960585e-05, "loss": 0.0337, "step": 113230 }, { "epoch": 2.2628991647016505, "grad_norm": 0.19513653218746185, "learning_rate": 2.1589762913353097e-05, "loss": 0.0289, "step": 113240 }, { "epoch": 2.2630989968426523, "grad_norm": 0.09894338250160217, "learning_rate": 2.158125965037876e-05, "loss": 0.0224, "step": 113250 }, { "epoch": 2.263298828983654, "grad_norm": 0.13917212188243866, "learning_rate": 2.157275760140083e-05, "loss": 0.0288, "step": 113260 }, { "epoch": 2.2634986611246553, "grad_norm": 0.12378598749637604, "learning_rate": 2.1564256766782443e-05, "loss": 0.0258, "step": 113270 }, { "epoch": 2.263698493265657, "grad_norm": 0.09734215587377548, "learning_rate": 2.155575714688676e-05, "loss": 0.0243, "step": 113280 }, { "epoch": 2.2638983254066583, "grad_norm": 0.17070364952087402, "learning_rate": 2.154725874207691e-05, "loss": 0.0256, "step": 113290 }, { "epoch": 2.26409815754766, "grad_norm": 0.09404496848583221, "learning_rate": 2.153876155271588e-05, "loss": 0.0212, "step": 113300 }, { "epoch": 2.2642979896886617, "grad_norm": 0.13614875078201294, "learning_rate": 2.153026557916667e-05, "loss": 0.0255, "step": 113310 }, { "epoch": 2.264497821829663, "grad_norm": 0.11059298366308212, "learning_rate": 2.152177082179222e-05, "loss": 0.0287, "step": 113320 }, { "epoch": 2.2646976539706647, "grad_norm": 0.1327492743730545, "learning_rate": 2.1513277280955397e-05, "loss": 0.0291, "step": 113330 }, { "epoch": 2.264897486111666, "grad_norm": 0.13865695893764496, "learning_rate": 2.150478495701904e-05, "loss": 0.0261, "step": 113340 }, { "epoch": 2.2650973182526677, "grad_norm": 0.16529282927513123, "learning_rate": 2.149629385034592e-05, "loss": 0.0266, "step": 113350 }, { "epoch": 2.2652971503936694, "grad_norm": 0.11267212778329849, "learning_rate": 2.1487803961298786e-05, "loss": 0.0225, "step": 113360 }, { "epoch": 2.2654969825346707, "grad_norm": 0.11979326605796814, "learning_rate": 2.1479315290240248e-05, "loss": 0.0202, "step": 113370 }, { "epoch": 2.2656968146756724, "grad_norm": 0.1028149425983429, "learning_rate": 2.1470827837532987e-05, "loss": 0.026, "step": 113380 }, { "epoch": 2.265896646816674, "grad_norm": 0.11803130805492401, "learning_rate": 2.1462341603539572e-05, "loss": 0.0168, "step": 113390 }, { "epoch": 2.2660964789576754, "grad_norm": 0.10749050229787827, "learning_rate": 2.1453856588622475e-05, "loss": 0.0275, "step": 113400 }, { "epoch": 2.266296311098677, "grad_norm": 0.14867539703845978, "learning_rate": 2.1445372793144186e-05, "loss": 0.0284, "step": 113410 }, { "epoch": 2.266496143239679, "grad_norm": 0.15543191134929657, "learning_rate": 2.1436890217467114e-05, "loss": 0.0276, "step": 113420 }, { "epoch": 2.26669597538068, "grad_norm": 0.14040152728557587, "learning_rate": 2.142840886195362e-05, "loss": 0.0244, "step": 113430 }, { "epoch": 2.266895807521682, "grad_norm": 0.13094203174114227, "learning_rate": 2.141992872696601e-05, "loss": 0.0296, "step": 113440 }, { "epoch": 2.267095639662683, "grad_norm": 0.15996317565441132, "learning_rate": 2.1411449812866546e-05, "loss": 0.0257, "step": 113450 }, { "epoch": 2.267295471803685, "grad_norm": 0.118330217897892, "learning_rate": 2.140297212001742e-05, "loss": 0.0243, "step": 113460 }, { "epoch": 2.2674953039446866, "grad_norm": 0.0939612165093422, "learning_rate": 2.13944956487808e-05, "loss": 0.022, "step": 113470 }, { "epoch": 2.267695136085688, "grad_norm": 0.13009750843048096, "learning_rate": 2.138602039951877e-05, "loss": 0.0188, "step": 113480 }, { "epoch": 2.2678949682266896, "grad_norm": 0.16367249190807343, "learning_rate": 2.1377546372593397e-05, "loss": 0.0268, "step": 113490 }, { "epoch": 2.2680948003676913, "grad_norm": 0.10735010355710983, "learning_rate": 2.1369073568366654e-05, "loss": 0.0232, "step": 113500 }, { "epoch": 2.2682946325086926, "grad_norm": 0.1245170310139656, "learning_rate": 2.1360601987200467e-05, "loss": 0.0305, "step": 113510 }, { "epoch": 2.2684944646496943, "grad_norm": 0.1355070173740387, "learning_rate": 2.1352131629456783e-05, "loss": 0.0247, "step": 113520 }, { "epoch": 2.268694296790696, "grad_norm": 0.11764131486415863, "learning_rate": 2.1343662495497398e-05, "loss": 0.0383, "step": 113530 }, { "epoch": 2.2688941289316973, "grad_norm": 0.16894735395908356, "learning_rate": 2.1335194585684105e-05, "loss": 0.0189, "step": 113540 }, { "epoch": 2.269093961072699, "grad_norm": 0.14936469495296478, "learning_rate": 2.1326727900378622e-05, "loss": 0.0307, "step": 113550 }, { "epoch": 2.2692937932137003, "grad_norm": 0.13864079117774963, "learning_rate": 2.1318262439942687e-05, "loss": 0.0251, "step": 113560 }, { "epoch": 2.269493625354702, "grad_norm": 0.14977964758872986, "learning_rate": 2.130979820473787e-05, "loss": 0.0225, "step": 113570 }, { "epoch": 2.2696934574957037, "grad_norm": 0.10497655719518661, "learning_rate": 2.1301335195125767e-05, "loss": 0.0215, "step": 113580 }, { "epoch": 2.269893289636705, "grad_norm": 0.1498195379972458, "learning_rate": 2.1292873411467908e-05, "loss": 0.021, "step": 113590 }, { "epoch": 2.2700931217777067, "grad_norm": 0.1038745865225792, "learning_rate": 2.1284412854125752e-05, "loss": 0.025, "step": 113600 }, { "epoch": 2.2702929539187084, "grad_norm": 0.10208628326654434, "learning_rate": 2.1275953523460733e-05, "loss": 0.0228, "step": 113610 }, { "epoch": 2.2704927860597097, "grad_norm": 0.1413378268480301, "learning_rate": 2.126749541983421e-05, "loss": 0.0305, "step": 113620 }, { "epoch": 2.2706926182007114, "grad_norm": 0.1471739560365677, "learning_rate": 2.125903854360752e-05, "loss": 0.0225, "step": 113630 }, { "epoch": 2.270892450341713, "grad_norm": 0.0941176488995552, "learning_rate": 2.1250582895141864e-05, "loss": 0.026, "step": 113640 }, { "epoch": 2.2710922824827144, "grad_norm": 0.12056652456521988, "learning_rate": 2.1242128474798512e-05, "loss": 0.0223, "step": 113650 }, { "epoch": 2.271292114623716, "grad_norm": 0.1672084480524063, "learning_rate": 2.123367528293862e-05, "loss": 0.0211, "step": 113660 }, { "epoch": 2.2714919467647174, "grad_norm": 0.11217833310365677, "learning_rate": 2.122522331992326e-05, "loss": 0.0271, "step": 113670 }, { "epoch": 2.271691778905719, "grad_norm": 0.1132248044013977, "learning_rate": 2.1216772586113477e-05, "loss": 0.0218, "step": 113680 }, { "epoch": 2.271891611046721, "grad_norm": 0.12233275175094604, "learning_rate": 2.120832308187033e-05, "loss": 0.0244, "step": 113690 }, { "epoch": 2.272091443187722, "grad_norm": 0.1130976602435112, "learning_rate": 2.119987480755471e-05, "loss": 0.0217, "step": 113700 }, { "epoch": 2.272291275328724, "grad_norm": 0.11237355321645737, "learning_rate": 2.1191427763527525e-05, "loss": 0.02, "step": 113710 }, { "epoch": 2.2724911074697256, "grad_norm": 0.1063990369439125, "learning_rate": 2.1182981950149632e-05, "loss": 0.0257, "step": 113720 }, { "epoch": 2.272690939610727, "grad_norm": 0.13289937376976013, "learning_rate": 2.117453736778181e-05, "loss": 0.0219, "step": 113730 }, { "epoch": 2.2728907717517286, "grad_norm": 0.1296195536851883, "learning_rate": 2.1166094016784792e-05, "loss": 0.0197, "step": 113740 }, { "epoch": 2.2730906038927303, "grad_norm": 0.12530525028705597, "learning_rate": 2.1157651897519277e-05, "loss": 0.0263, "step": 113750 }, { "epoch": 2.2732904360337316, "grad_norm": 0.13353963196277618, "learning_rate": 2.11492110103459e-05, "loss": 0.0263, "step": 113760 }, { "epoch": 2.2734902681747333, "grad_norm": 0.13526400923728943, "learning_rate": 2.1140771355625194e-05, "loss": 0.0236, "step": 113770 }, { "epoch": 2.2736901003157346, "grad_norm": 0.16427285969257355, "learning_rate": 2.1132332933717735e-05, "loss": 0.0302, "step": 113780 }, { "epoch": 2.2738899324567363, "grad_norm": 0.1192585751414299, "learning_rate": 2.1123895744984003e-05, "loss": 0.0262, "step": 113790 }, { "epoch": 2.274089764597738, "grad_norm": 0.12606969475746155, "learning_rate": 2.1115459789784375e-05, "loss": 0.0222, "step": 113800 }, { "epoch": 2.2742895967387393, "grad_norm": 0.11268577724695206, "learning_rate": 2.110702506847924e-05, "loss": 0.0245, "step": 113810 }, { "epoch": 2.274489428879741, "grad_norm": 0.10622791200876236, "learning_rate": 2.109859158142892e-05, "loss": 0.0246, "step": 113820 }, { "epoch": 2.2746892610207428, "grad_norm": 0.11307227611541748, "learning_rate": 2.109015932899368e-05, "loss": 0.0251, "step": 113830 }, { "epoch": 2.274889093161744, "grad_norm": 0.10761462897062302, "learning_rate": 2.1081728311533715e-05, "loss": 0.0191, "step": 113840 }, { "epoch": 2.2750889253027458, "grad_norm": 0.10293301194906235, "learning_rate": 2.1073298529409204e-05, "loss": 0.0251, "step": 113850 }, { "epoch": 2.2752887574437475, "grad_norm": 0.11816968023777008, "learning_rate": 2.106486998298023e-05, "loss": 0.0311, "step": 113860 }, { "epoch": 2.2754885895847488, "grad_norm": 0.12488199770450592, "learning_rate": 2.1056442672606853e-05, "loss": 0.0267, "step": 113870 }, { "epoch": 2.2756884217257505, "grad_norm": 0.12383835762739182, "learning_rate": 2.104801659864908e-05, "loss": 0.0191, "step": 113880 }, { "epoch": 2.2758882538667518, "grad_norm": 0.10016509145498276, "learning_rate": 2.1039591761466864e-05, "loss": 0.0271, "step": 113890 }, { "epoch": 2.2760880860077535, "grad_norm": 0.11545241624116898, "learning_rate": 2.1031168161420074e-05, "loss": 0.0325, "step": 113900 }, { "epoch": 2.276287918148755, "grad_norm": 0.10068660229444504, "learning_rate": 2.1022745798868544e-05, "loss": 0.0215, "step": 113910 }, { "epoch": 2.2764877502897565, "grad_norm": 0.14434240758419037, "learning_rate": 2.101432467417211e-05, "loss": 0.0281, "step": 113920 }, { "epoch": 2.276687582430758, "grad_norm": 0.17783887684345245, "learning_rate": 2.1005904787690465e-05, "loss": 0.0283, "step": 113930 }, { "epoch": 2.27688741457176, "grad_norm": 0.10904394090175629, "learning_rate": 2.09974861397833e-05, "loss": 0.0265, "step": 113940 }, { "epoch": 2.277087246712761, "grad_norm": 0.12693382799625397, "learning_rate": 2.0989068730810236e-05, "loss": 0.0168, "step": 113950 }, { "epoch": 2.277287078853763, "grad_norm": 0.17492379248142242, "learning_rate": 2.0980652561130894e-05, "loss": 0.0232, "step": 113960 }, { "epoch": 2.2774869109947646, "grad_norm": 0.0969131588935852, "learning_rate": 2.0972237631104747e-05, "loss": 0.0242, "step": 113970 }, { "epoch": 2.277686743135766, "grad_norm": 0.13844607770442963, "learning_rate": 2.0963823941091277e-05, "loss": 0.0304, "step": 113980 }, { "epoch": 2.2778865752767676, "grad_norm": 0.12392154335975647, "learning_rate": 2.095541149144991e-05, "loss": 0.0335, "step": 113990 }, { "epoch": 2.278086407417769, "grad_norm": 0.15099047124385834, "learning_rate": 2.0947000282540014e-05, "loss": 0.0312, "step": 114000 }, { "epoch": 2.2782862395587706, "grad_norm": 0.12521420419216156, "learning_rate": 2.0938590314720884e-05, "loss": 0.027, "step": 114010 }, { "epoch": 2.2784860716997724, "grad_norm": 0.10344844311475754, "learning_rate": 2.0930181588351795e-05, "loss": 0.0246, "step": 114020 }, { "epoch": 2.2786859038407736, "grad_norm": 0.18062922358512878, "learning_rate": 2.0921774103791964e-05, "loss": 0.0278, "step": 114030 }, { "epoch": 2.2788857359817754, "grad_norm": 0.08982428908348083, "learning_rate": 2.0913367861400485e-05, "loss": 0.0212, "step": 114040 }, { "epoch": 2.2790855681227766, "grad_norm": 0.13525055348873138, "learning_rate": 2.0904962861536516e-05, "loss": 0.0254, "step": 114050 }, { "epoch": 2.2792854002637783, "grad_norm": 0.1549917757511139, "learning_rate": 2.0896559104559106e-05, "loss": 0.0292, "step": 114060 }, { "epoch": 2.27948523240478, "grad_norm": 0.13296227157115936, "learning_rate": 2.0888156590827203e-05, "loss": 0.0221, "step": 114070 }, { "epoch": 2.279685064545782, "grad_norm": 0.14070583879947662, "learning_rate": 2.0879755320699772e-05, "loss": 0.0159, "step": 114080 }, { "epoch": 2.279884896686783, "grad_norm": 0.14520163834095, "learning_rate": 2.08713552945357e-05, "loss": 0.0239, "step": 114090 }, { "epoch": 2.280084728827785, "grad_norm": 0.11204439401626587, "learning_rate": 2.0862956512693816e-05, "loss": 0.023, "step": 114100 }, { "epoch": 2.280284560968786, "grad_norm": 0.09963268041610718, "learning_rate": 2.0854558975532908e-05, "loss": 0.0262, "step": 114110 }, { "epoch": 2.280484393109788, "grad_norm": 0.12058831751346588, "learning_rate": 2.084616268341169e-05, "loss": 0.0198, "step": 114120 }, { "epoch": 2.2806842252507895, "grad_norm": 0.1387958973646164, "learning_rate": 2.0837767636688854e-05, "loss": 0.0222, "step": 114130 }, { "epoch": 2.280884057391791, "grad_norm": 0.20796585083007812, "learning_rate": 2.0829373835723005e-05, "loss": 0.0268, "step": 114140 }, { "epoch": 2.2810838895327925, "grad_norm": 0.13446567952632904, "learning_rate": 2.082098128087272e-05, "loss": 0.0276, "step": 114150 }, { "epoch": 2.281283721673794, "grad_norm": 0.09784970432519913, "learning_rate": 2.0812589972496532e-05, "loss": 0.0209, "step": 114160 }, { "epoch": 2.2814835538147955, "grad_norm": 0.1088537722826004, "learning_rate": 2.080419991095286e-05, "loss": 0.0289, "step": 114170 }, { "epoch": 2.2816833859557972, "grad_norm": 0.21586455404758453, "learning_rate": 2.0795811096600115e-05, "loss": 0.0244, "step": 114180 }, { "epoch": 2.2818832180967985, "grad_norm": 0.1122688427567482, "learning_rate": 2.0787423529796706e-05, "loss": 0.0208, "step": 114190 }, { "epoch": 2.2820830502378002, "grad_norm": 0.1307549923658371, "learning_rate": 2.0779037210900887e-05, "loss": 0.0193, "step": 114200 }, { "epoch": 2.282282882378802, "grad_norm": 0.11011544615030289, "learning_rate": 2.0770652140270912e-05, "loss": 0.0238, "step": 114210 }, { "epoch": 2.2824827145198032, "grad_norm": 0.11645511537790298, "learning_rate": 2.076226831826498e-05, "loss": 0.027, "step": 114220 }, { "epoch": 2.282682546660805, "grad_norm": 0.11099149286746979, "learning_rate": 2.0753885745241237e-05, "loss": 0.0236, "step": 114230 }, { "epoch": 2.2828823788018067, "grad_norm": 0.12080509215593338, "learning_rate": 2.0745504421557772e-05, "loss": 0.026, "step": 114240 }, { "epoch": 2.283082210942808, "grad_norm": 0.13921187818050385, "learning_rate": 2.0737124347572605e-05, "loss": 0.0277, "step": 114250 }, { "epoch": 2.2832820430838097, "grad_norm": 0.10655257850885391, "learning_rate": 2.0728745523643757e-05, "loss": 0.0329, "step": 114260 }, { "epoch": 2.283481875224811, "grad_norm": 0.13758616149425507, "learning_rate": 2.0720367950129087e-05, "loss": 0.0245, "step": 114270 }, { "epoch": 2.2836817073658127, "grad_norm": 0.13987529277801514, "learning_rate": 2.071199162738653e-05, "loss": 0.0263, "step": 114280 }, { "epoch": 2.2838815395068144, "grad_norm": 0.08403939008712769, "learning_rate": 2.0703616555773895e-05, "loss": 0.0296, "step": 114290 }, { "epoch": 2.2840813716478157, "grad_norm": 0.15116490423679352, "learning_rate": 2.0695242735648934e-05, "loss": 0.0297, "step": 114300 }, { "epoch": 2.2842812037888174, "grad_norm": 0.16609390079975128, "learning_rate": 2.0686870167369348e-05, "loss": 0.0266, "step": 114310 }, { "epoch": 2.284481035929819, "grad_norm": 0.1417253464460373, "learning_rate": 2.0678498851292855e-05, "loss": 0.0332, "step": 114320 }, { "epoch": 2.2846808680708204, "grad_norm": 0.0998309999704361, "learning_rate": 2.067012878777701e-05, "loss": 0.0188, "step": 114330 }, { "epoch": 2.284880700211822, "grad_norm": 0.091835156083107, "learning_rate": 2.066175997717938e-05, "loss": 0.0214, "step": 114340 }, { "epoch": 2.285080532352824, "grad_norm": 0.1445552259683609, "learning_rate": 2.0653392419857463e-05, "loss": 0.0276, "step": 114350 }, { "epoch": 2.285280364493825, "grad_norm": 0.13843724131584167, "learning_rate": 2.064502611616872e-05, "loss": 0.0275, "step": 114360 }, { "epoch": 2.285480196634827, "grad_norm": 0.12544776499271393, "learning_rate": 2.0636661066470526e-05, "loss": 0.0273, "step": 114370 }, { "epoch": 2.285680028775828, "grad_norm": 0.1189543828368187, "learning_rate": 2.0628297271120232e-05, "loss": 0.0209, "step": 114380 }, { "epoch": 2.28587986091683, "grad_norm": 0.12048763036727905, "learning_rate": 2.061993473047512e-05, "loss": 0.0286, "step": 114390 }, { "epoch": 2.2860796930578315, "grad_norm": 0.11887692660093307, "learning_rate": 2.0611573444892428e-05, "loss": 0.0223, "step": 114400 }, { "epoch": 2.286279525198833, "grad_norm": 0.11670324206352234, "learning_rate": 2.060321341472933e-05, "loss": 0.0206, "step": 114410 }, { "epoch": 2.2864793573398345, "grad_norm": 0.09948592633008957, "learning_rate": 2.059485464034294e-05, "loss": 0.0231, "step": 114420 }, { "epoch": 2.2866791894808363, "grad_norm": 0.14886459708213806, "learning_rate": 2.058649712209037e-05, "loss": 0.028, "step": 114430 }, { "epoch": 2.2868790216218375, "grad_norm": 0.10463125258684158, "learning_rate": 2.0578140860328588e-05, "loss": 0.0217, "step": 114440 }, { "epoch": 2.2870788537628393, "grad_norm": 0.13099662959575653, "learning_rate": 2.0569785855414565e-05, "loss": 0.0199, "step": 114450 }, { "epoch": 2.287278685903841, "grad_norm": 0.17085225880146027, "learning_rate": 2.0561432107705254e-05, "loss": 0.0259, "step": 114460 }, { "epoch": 2.2874785180448423, "grad_norm": 0.11259432137012482, "learning_rate": 2.055307961755747e-05, "loss": 0.0314, "step": 114470 }, { "epoch": 2.287678350185844, "grad_norm": 0.29184117913246155, "learning_rate": 2.054472838532802e-05, "loss": 0.0283, "step": 114480 }, { "epoch": 2.2878781823268453, "grad_norm": 0.0989319458603859, "learning_rate": 2.0536378411373676e-05, "loss": 0.0276, "step": 114490 }, { "epoch": 2.288078014467847, "grad_norm": 0.11930612474679947, "learning_rate": 2.052802969605111e-05, "loss": 0.0268, "step": 114500 }, { "epoch": 2.2882778466088487, "grad_norm": 0.12983456254005432, "learning_rate": 2.0519682239716975e-05, "loss": 0.0279, "step": 114510 }, { "epoch": 2.28847767874985, "grad_norm": 0.128037229180336, "learning_rate": 2.0511336042727862e-05, "loss": 0.0237, "step": 114520 }, { "epoch": 2.2886775108908517, "grad_norm": 0.1132199689745903, "learning_rate": 2.0502991105440317e-05, "loss": 0.0254, "step": 114530 }, { "epoch": 2.2888773430318534, "grad_norm": 0.10765454173088074, "learning_rate": 2.0494647428210768e-05, "loss": 0.0184, "step": 114540 }, { "epoch": 2.2890771751728547, "grad_norm": 0.09638427197933197, "learning_rate": 2.048630501139569e-05, "loss": 0.0317, "step": 114550 }, { "epoch": 2.2892770073138564, "grad_norm": 0.09384972602128983, "learning_rate": 2.0477963855351466e-05, "loss": 0.0255, "step": 114560 }, { "epoch": 2.289476839454858, "grad_norm": 0.12213335186243057, "learning_rate": 2.0469623960434376e-05, "loss": 0.0246, "step": 114570 }, { "epoch": 2.2896766715958594, "grad_norm": 0.11989960819482803, "learning_rate": 2.0461285327000683e-05, "loss": 0.0239, "step": 114580 }, { "epoch": 2.289876503736861, "grad_norm": 0.14452840387821198, "learning_rate": 2.0452947955406653e-05, "loss": 0.024, "step": 114590 }, { "epoch": 2.2900763358778624, "grad_norm": 0.12747976183891296, "learning_rate": 2.0444611846008387e-05, "loss": 0.03, "step": 114600 }, { "epoch": 2.290276168018864, "grad_norm": 0.09945927560329437, "learning_rate": 2.0436276999162006e-05, "loss": 0.0264, "step": 114610 }, { "epoch": 2.290476000159866, "grad_norm": 0.08839816600084305, "learning_rate": 2.0427943415223566e-05, "loss": 0.0249, "step": 114620 }, { "epoch": 2.290675832300867, "grad_norm": 0.15966260433197021, "learning_rate": 2.0419611094549057e-05, "loss": 0.0278, "step": 114630 }, { "epoch": 2.290875664441869, "grad_norm": 0.15392005443572998, "learning_rate": 2.041128003749442e-05, "loss": 0.0189, "step": 114640 }, { "epoch": 2.2910754965828706, "grad_norm": 0.09225614368915558, "learning_rate": 2.0402950244415542e-05, "loss": 0.0252, "step": 114650 }, { "epoch": 2.291275328723872, "grad_norm": 0.11749496310949326, "learning_rate": 2.0394621715668278e-05, "loss": 0.0299, "step": 114660 }, { "epoch": 2.2914751608648736, "grad_norm": 0.11137283593416214, "learning_rate": 2.038629445160835e-05, "loss": 0.0384, "step": 114670 }, { "epoch": 2.2916749930058753, "grad_norm": 0.12262780219316483, "learning_rate": 2.037796845259154e-05, "loss": 0.0282, "step": 114680 }, { "epoch": 2.2918748251468766, "grad_norm": 0.12284388393163681, "learning_rate": 2.036964371897352e-05, "loss": 0.0208, "step": 114690 }, { "epoch": 2.2920746572878783, "grad_norm": 0.12421870231628418, "learning_rate": 2.0361320251109865e-05, "loss": 0.0214, "step": 114700 }, { "epoch": 2.2922744894288796, "grad_norm": 0.10646998137235641, "learning_rate": 2.0352998049356164e-05, "loss": 0.022, "step": 114710 }, { "epoch": 2.2924743215698813, "grad_norm": 0.11028063297271729, "learning_rate": 2.0344677114067924e-05, "loss": 0.0186, "step": 114720 }, { "epoch": 2.292674153710883, "grad_norm": 0.11870156228542328, "learning_rate": 2.0336357445600594e-05, "loss": 0.0319, "step": 114730 }, { "epoch": 2.2928739858518843, "grad_norm": 0.12877415120601654, "learning_rate": 2.0328039044309584e-05, "loss": 0.0276, "step": 114740 }, { "epoch": 2.293073817992886, "grad_norm": 0.15468457341194153, "learning_rate": 2.031972191055024e-05, "loss": 0.0238, "step": 114750 }, { "epoch": 2.2932736501338873, "grad_norm": 0.13213057816028595, "learning_rate": 2.0311406044677844e-05, "loss": 0.0199, "step": 114760 }, { "epoch": 2.293473482274889, "grad_norm": 0.14713823795318604, "learning_rate": 2.0303091447047646e-05, "loss": 0.0228, "step": 114770 }, { "epoch": 2.2936733144158907, "grad_norm": 0.11958760023117065, "learning_rate": 2.0294778118014828e-05, "loss": 0.0297, "step": 114780 }, { "epoch": 2.2938731465568925, "grad_norm": 0.12090179324150085, "learning_rate": 2.0286466057934516e-05, "loss": 0.0241, "step": 114790 }, { "epoch": 2.2940729786978937, "grad_norm": 0.11102984100580215, "learning_rate": 2.027815526716179e-05, "loss": 0.0265, "step": 114800 }, { "epoch": 2.2942728108388954, "grad_norm": 0.08355245739221573, "learning_rate": 2.0269845746051668e-05, "loss": 0.0244, "step": 114810 }, { "epoch": 2.2944726429798967, "grad_norm": 0.20018739998340607, "learning_rate": 2.0261537494959126e-05, "loss": 0.0277, "step": 114820 }, { "epoch": 2.2946724751208984, "grad_norm": 0.09946445375680923, "learning_rate": 2.025323051423908e-05, "loss": 0.0217, "step": 114830 }, { "epoch": 2.2948723072619, "grad_norm": 0.15462201833724976, "learning_rate": 2.0244924804246368e-05, "loss": 0.0277, "step": 114840 }, { "epoch": 2.2950721394029014, "grad_norm": 0.1402726024389267, "learning_rate": 2.023662036533579e-05, "loss": 0.0289, "step": 114850 }, { "epoch": 2.295271971543903, "grad_norm": 0.14971545338630676, "learning_rate": 2.0228317197862147e-05, "loss": 0.0198, "step": 114860 }, { "epoch": 2.2954718036849044, "grad_norm": 0.16342082619667053, "learning_rate": 2.022001530218008e-05, "loss": 0.0284, "step": 114870 }, { "epoch": 2.295671635825906, "grad_norm": 0.13873495161533356, "learning_rate": 2.021171467864425e-05, "loss": 0.0247, "step": 114880 }, { "epoch": 2.295871467966908, "grad_norm": 0.12346795946359634, "learning_rate": 2.020341532760925e-05, "loss": 0.0217, "step": 114890 }, { "epoch": 2.296071300107909, "grad_norm": 0.08557292073965073, "learning_rate": 2.0195117249429606e-05, "loss": 0.0271, "step": 114900 }, { "epoch": 2.296271132248911, "grad_norm": 0.14673511683940887, "learning_rate": 2.018682044445981e-05, "loss": 0.0235, "step": 114910 }, { "epoch": 2.2964709643899126, "grad_norm": 0.1164805069565773, "learning_rate": 2.0178524913054265e-05, "loss": 0.021, "step": 114920 }, { "epoch": 2.296670796530914, "grad_norm": 0.14831934869289398, "learning_rate": 2.0170230655567374e-05, "loss": 0.0263, "step": 114930 }, { "epoch": 2.2968706286719156, "grad_norm": 0.1297018676996231, "learning_rate": 2.0161937672353393e-05, "loss": 0.029, "step": 114940 }, { "epoch": 2.2970704608129173, "grad_norm": 0.13508176803588867, "learning_rate": 2.0153645963766636e-05, "loss": 0.0214, "step": 114950 }, { "epoch": 2.2972702929539186, "grad_norm": 0.15372662246227264, "learning_rate": 2.014535553016132e-05, "loss": 0.0232, "step": 114960 }, { "epoch": 2.2974701250949203, "grad_norm": 0.10560941696166992, "learning_rate": 2.0137066371891544e-05, "loss": 0.0275, "step": 114970 }, { "epoch": 2.2976699572359216, "grad_norm": 0.15808182954788208, "learning_rate": 2.0128778489311424e-05, "loss": 0.0286, "step": 114980 }, { "epoch": 2.2978697893769233, "grad_norm": 0.13934452831745148, "learning_rate": 2.0120491882775038e-05, "loss": 0.0287, "step": 114990 }, { "epoch": 2.298069621517925, "grad_norm": 0.1256549209356308, "learning_rate": 2.0112206552636336e-05, "loss": 0.0317, "step": 115000 }, { "epoch": 2.2982694536589263, "grad_norm": 0.10079961270093918, "learning_rate": 2.0103922499249262e-05, "loss": 0.0332, "step": 115010 }, { "epoch": 2.298469285799928, "grad_norm": 0.12380611896514893, "learning_rate": 2.00956397229677e-05, "loss": 0.027, "step": 115020 }, { "epoch": 2.2986691179409298, "grad_norm": 0.17789623141288757, "learning_rate": 2.0087358224145476e-05, "loss": 0.0236, "step": 115030 }, { "epoch": 2.298868950081931, "grad_norm": 0.1792793869972229, "learning_rate": 2.0079078003136355e-05, "loss": 0.0315, "step": 115040 }, { "epoch": 2.2990687822229328, "grad_norm": 0.14001627266407013, "learning_rate": 2.007079906029405e-05, "loss": 0.0278, "step": 115050 }, { "epoch": 2.2992686143639345, "grad_norm": 0.11519809812307358, "learning_rate": 2.006252139597225e-05, "loss": 0.0212, "step": 115060 }, { "epoch": 2.2994684465049358, "grad_norm": 0.17485831677913666, "learning_rate": 2.00542450105245e-05, "loss": 0.0283, "step": 115070 }, { "epoch": 2.2996682786459375, "grad_norm": 0.1343979686498642, "learning_rate": 2.0045969904304412e-05, "loss": 0.0259, "step": 115080 }, { "epoch": 2.2998681107869388, "grad_norm": 0.15321402251720428, "learning_rate": 2.003769607766548e-05, "loss": 0.0296, "step": 115090 }, { "epoch": 2.3000679429279405, "grad_norm": 0.1493777632713318, "learning_rate": 2.002942353096111e-05, "loss": 0.0227, "step": 115100 }, { "epoch": 2.300267775068942, "grad_norm": 0.12743230164051056, "learning_rate": 2.0021152264544713e-05, "loss": 0.0358, "step": 115110 }, { "epoch": 2.3004676072099435, "grad_norm": 0.15818297863006592, "learning_rate": 2.0012882278769612e-05, "loss": 0.025, "step": 115120 }, { "epoch": 2.300667439350945, "grad_norm": 0.137570321559906, "learning_rate": 2.0004613573989095e-05, "loss": 0.0228, "step": 115130 }, { "epoch": 2.300867271491947, "grad_norm": 0.13182000815868378, "learning_rate": 1.9996346150556382e-05, "loss": 0.0315, "step": 115140 }, { "epoch": 2.301067103632948, "grad_norm": 0.11476162821054459, "learning_rate": 1.9988080008824645e-05, "loss": 0.0213, "step": 115150 }, { "epoch": 2.30126693577395, "grad_norm": 0.13200359046459198, "learning_rate": 1.9979815149146997e-05, "loss": 0.021, "step": 115160 }, { "epoch": 2.3014667679149516, "grad_norm": 0.14732137322425842, "learning_rate": 1.9971551571876496e-05, "loss": 0.0259, "step": 115170 }, { "epoch": 2.301666600055953, "grad_norm": 0.09094057232141495, "learning_rate": 1.996328927736615e-05, "loss": 0.0273, "step": 115180 }, { "epoch": 2.3018664321969546, "grad_norm": 0.12154047191143036, "learning_rate": 1.99550282659689e-05, "loss": 0.0237, "step": 115190 }, { "epoch": 2.302066264337956, "grad_norm": 0.2262859344482422, "learning_rate": 1.994676853803767e-05, "loss": 0.0232, "step": 115200 }, { "epoch": 2.3022660964789576, "grad_norm": 0.10893889516592026, "learning_rate": 1.9938510093925244e-05, "loss": 0.0258, "step": 115210 }, { "epoch": 2.3024659286199594, "grad_norm": 0.12498533725738525, "learning_rate": 1.9930252933984457e-05, "loss": 0.0325, "step": 115220 }, { "epoch": 2.3026657607609606, "grad_norm": 0.11098005622625351, "learning_rate": 1.9921997058568044e-05, "loss": 0.0309, "step": 115230 }, { "epoch": 2.3028655929019624, "grad_norm": 0.15725764632225037, "learning_rate": 1.9913742468028645e-05, "loss": 0.0225, "step": 115240 }, { "epoch": 2.303065425042964, "grad_norm": 0.1586802750825882, "learning_rate": 1.9905489162718876e-05, "loss": 0.0313, "step": 115250 }, { "epoch": 2.3032652571839654, "grad_norm": 0.13523969054222107, "learning_rate": 1.9897237142991367e-05, "loss": 0.0206, "step": 115260 }, { "epoch": 2.303465089324967, "grad_norm": 0.14274661242961884, "learning_rate": 1.988898640919856e-05, "loss": 0.0331, "step": 115270 }, { "epoch": 2.303664921465969, "grad_norm": 0.15940868854522705, "learning_rate": 1.9880736961692947e-05, "loss": 0.0257, "step": 115280 }, { "epoch": 2.30386475360697, "grad_norm": 0.1512930989265442, "learning_rate": 1.9872488800826918e-05, "loss": 0.0284, "step": 115290 }, { "epoch": 2.304064585747972, "grad_norm": 0.12233179807662964, "learning_rate": 1.9864241926952814e-05, "loss": 0.0304, "step": 115300 }, { "epoch": 2.304264417888973, "grad_norm": 0.16570763289928436, "learning_rate": 1.9855996340422945e-05, "loss": 0.0205, "step": 115310 }, { "epoch": 2.304464250029975, "grad_norm": 0.13411246240139008, "learning_rate": 1.984775204158953e-05, "loss": 0.0243, "step": 115320 }, { "epoch": 2.3046640821709765, "grad_norm": 0.1439645141363144, "learning_rate": 1.983950903080477e-05, "loss": 0.0276, "step": 115330 }, { "epoch": 2.304863914311978, "grad_norm": 0.09045839309692383, "learning_rate": 1.9831267308420746e-05, "loss": 0.0239, "step": 115340 }, { "epoch": 2.3050637464529795, "grad_norm": 0.12751813232898712, "learning_rate": 1.982302687478958e-05, "loss": 0.0284, "step": 115350 }, { "epoch": 2.3052635785939812, "grad_norm": 0.10558845102787018, "learning_rate": 1.9814787730263286e-05, "loss": 0.0269, "step": 115360 }, { "epoch": 2.3054634107349825, "grad_norm": 0.10889551788568497, "learning_rate": 1.9806549875193792e-05, "loss": 0.0342, "step": 115370 }, { "epoch": 2.3056632428759842, "grad_norm": 0.119852215051651, "learning_rate": 1.979831330993302e-05, "loss": 0.0264, "step": 115380 }, { "epoch": 2.305863075016986, "grad_norm": 0.2778594493865967, "learning_rate": 1.9790078034832816e-05, "loss": 0.0249, "step": 115390 }, { "epoch": 2.3060629071579872, "grad_norm": 0.11827471107244492, "learning_rate": 1.9781844050244986e-05, "loss": 0.026, "step": 115400 }, { "epoch": 2.306262739298989, "grad_norm": 0.08857796341180801, "learning_rate": 1.9773611356521267e-05, "loss": 0.025, "step": 115410 }, { "epoch": 2.3064625714399902, "grad_norm": 0.1465328335762024, "learning_rate": 1.9765379954013335e-05, "loss": 0.037, "step": 115420 }, { "epoch": 2.306662403580992, "grad_norm": 0.14355649054050446, "learning_rate": 1.9757149843072836e-05, "loss": 0.0256, "step": 115430 }, { "epoch": 2.3068622357219937, "grad_norm": 0.10555823892354965, "learning_rate": 1.9748921024051335e-05, "loss": 0.0198, "step": 115440 }, { "epoch": 2.307062067862995, "grad_norm": 0.1407238095998764, "learning_rate": 1.9740693497300355e-05, "loss": 0.0277, "step": 115450 }, { "epoch": 2.3072619000039967, "grad_norm": 0.1106920838356018, "learning_rate": 1.973246726317138e-05, "loss": 0.0309, "step": 115460 }, { "epoch": 2.3074617321449984, "grad_norm": 0.1391761600971222, "learning_rate": 1.972424232201578e-05, "loss": 0.0337, "step": 115470 }, { "epoch": 2.3076615642859997, "grad_norm": 0.2481801062822342, "learning_rate": 1.9716018674184915e-05, "loss": 0.0344, "step": 115480 }, { "epoch": 2.3078613964270014, "grad_norm": 0.08378869295120239, "learning_rate": 1.9707796320030132e-05, "loss": 0.0281, "step": 115490 }, { "epoch": 2.308061228568003, "grad_norm": 0.10351046174764633, "learning_rate": 1.9699575259902626e-05, "loss": 0.02, "step": 115500 }, { "epoch": 2.3082610607090044, "grad_norm": 0.13914012908935547, "learning_rate": 1.9691355494153596e-05, "loss": 0.0341, "step": 115510 }, { "epoch": 2.308460892850006, "grad_norm": 0.11993878334760666, "learning_rate": 1.9683137023134183e-05, "loss": 0.0328, "step": 115520 }, { "epoch": 2.3086607249910074, "grad_norm": 0.12250102311372757, "learning_rate": 1.9674919847195468e-05, "loss": 0.0196, "step": 115530 }, { "epoch": 2.308860557132009, "grad_norm": 0.13264861702919006, "learning_rate": 1.9666703966688464e-05, "loss": 0.0195, "step": 115540 }, { "epoch": 2.309060389273011, "grad_norm": 0.11057056486606598, "learning_rate": 1.965848938196414e-05, "loss": 0.0274, "step": 115550 }, { "epoch": 2.309260221414012, "grad_norm": 0.11709913611412048, "learning_rate": 1.965027609337343e-05, "loss": 0.0254, "step": 115560 }, { "epoch": 2.309460053555014, "grad_norm": 0.46971437335014343, "learning_rate": 1.9642064101267137e-05, "loss": 0.0296, "step": 115570 }, { "epoch": 2.309659885696015, "grad_norm": 0.07501482963562012, "learning_rate": 1.9633853405996115e-05, "loss": 0.0228, "step": 115580 }, { "epoch": 2.309859717837017, "grad_norm": 0.11667536944150925, "learning_rate": 1.9625644007911085e-05, "loss": 0.0273, "step": 115590 }, { "epoch": 2.3100595499780185, "grad_norm": 0.11822548508644104, "learning_rate": 1.961743590736277e-05, "loss": 0.0297, "step": 115600 }, { "epoch": 2.3102593821190203, "grad_norm": 0.12695135176181793, "learning_rate": 1.9609229104701737e-05, "loss": 0.0286, "step": 115610 }, { "epoch": 2.3104592142600215, "grad_norm": 0.10569049417972565, "learning_rate": 1.960102360027863e-05, "loss": 0.0256, "step": 115620 }, { "epoch": 2.3106590464010233, "grad_norm": 0.09381779283285141, "learning_rate": 1.9592819394443964e-05, "loss": 0.0297, "step": 115630 }, { "epoch": 2.3108588785420245, "grad_norm": 0.13022218644618988, "learning_rate": 1.9584616487548172e-05, "loss": 0.0272, "step": 115640 }, { "epoch": 2.3110587106830263, "grad_norm": 0.14582636952400208, "learning_rate": 1.9576414879941695e-05, "loss": 0.0283, "step": 115650 }, { "epoch": 2.311258542824028, "grad_norm": 0.1329285353422165, "learning_rate": 1.9568214571974886e-05, "loss": 0.0231, "step": 115660 }, { "epoch": 2.3114583749650293, "grad_norm": 0.12637637555599213, "learning_rate": 1.956001556399804e-05, "loss": 0.0256, "step": 115670 }, { "epoch": 2.311658207106031, "grad_norm": 0.11799774318933487, "learning_rate": 1.955181785636141e-05, "loss": 0.0218, "step": 115680 }, { "epoch": 2.3118580392470323, "grad_norm": 0.1693049818277359, "learning_rate": 1.9543621449415188e-05, "loss": 0.0225, "step": 115690 }, { "epoch": 2.312057871388034, "grad_norm": 0.09437688440084457, "learning_rate": 1.9535426343509506e-05, "loss": 0.0274, "step": 115700 }, { "epoch": 2.3122577035290357, "grad_norm": 0.10962702333927155, "learning_rate": 1.9527232538994443e-05, "loss": 0.0264, "step": 115710 }, { "epoch": 2.312457535670037, "grad_norm": 0.09258002787828445, "learning_rate": 1.9519040036220034e-05, "loss": 0.0282, "step": 115720 }, { "epoch": 2.3126573678110387, "grad_norm": 0.08655373752117157, "learning_rate": 1.9510848835536248e-05, "loss": 0.0252, "step": 115730 }, { "epoch": 2.3128571999520404, "grad_norm": 0.11763894557952881, "learning_rate": 1.950265893729298e-05, "loss": 0.0264, "step": 115740 }, { "epoch": 2.3130570320930417, "grad_norm": 0.14620210230350494, "learning_rate": 1.949447034184007e-05, "loss": 0.022, "step": 115750 }, { "epoch": 2.3132568642340434, "grad_norm": 0.11815983057022095, "learning_rate": 1.948628304952739e-05, "loss": 0.0255, "step": 115760 }, { "epoch": 2.313456696375045, "grad_norm": 0.10986652225255966, "learning_rate": 1.947809706070462e-05, "loss": 0.0318, "step": 115770 }, { "epoch": 2.3136565285160464, "grad_norm": 0.11652952432632446, "learning_rate": 1.946991237572147e-05, "loss": 0.0244, "step": 115780 }, { "epoch": 2.313856360657048, "grad_norm": 0.12034597247838974, "learning_rate": 1.9461728994927587e-05, "loss": 0.0299, "step": 115790 }, { "epoch": 2.3140561927980494, "grad_norm": 0.22673524916172028, "learning_rate": 1.9453546918672533e-05, "loss": 0.0313, "step": 115800 }, { "epoch": 2.314256024939051, "grad_norm": 0.13832798600196838, "learning_rate": 1.9445366147305842e-05, "loss": 0.0318, "step": 115810 }, { "epoch": 2.314455857080053, "grad_norm": 0.12483133375644684, "learning_rate": 1.9437186681176977e-05, "loss": 0.0248, "step": 115820 }, { "epoch": 2.314655689221054, "grad_norm": 0.1220056340098381, "learning_rate": 1.9429008520635372e-05, "loss": 0.0248, "step": 115830 }, { "epoch": 2.314855521362056, "grad_norm": 0.15165849030017853, "learning_rate": 1.9420831666030325e-05, "loss": 0.0316, "step": 115840 }, { "epoch": 2.3150553535030576, "grad_norm": 0.09330357611179352, "learning_rate": 1.941265611771119e-05, "loss": 0.0298, "step": 115850 }, { "epoch": 2.315255185644059, "grad_norm": 0.09782309085130692, "learning_rate": 1.940448187602722e-05, "loss": 0.0261, "step": 115860 }, { "epoch": 2.3154550177850606, "grad_norm": 0.14061540365219116, "learning_rate": 1.939630894132755e-05, "loss": 0.0261, "step": 115870 }, { "epoch": 2.3156548499260623, "grad_norm": 0.10787500441074371, "learning_rate": 1.938813731396133e-05, "loss": 0.0256, "step": 115880 }, { "epoch": 2.3158546820670636, "grad_norm": 0.09458466619253159, "learning_rate": 1.9379966994277683e-05, "loss": 0.0229, "step": 115890 }, { "epoch": 2.3160545142080653, "grad_norm": 0.12917885184288025, "learning_rate": 1.9371797982625576e-05, "loss": 0.0294, "step": 115900 }, { "epoch": 2.3162543463490666, "grad_norm": 0.11215522885322571, "learning_rate": 1.9363630279353994e-05, "loss": 0.0289, "step": 115910 }, { "epoch": 2.3164541784900683, "grad_norm": 0.14180324971675873, "learning_rate": 1.9355463884811847e-05, "loss": 0.0256, "step": 115920 }, { "epoch": 2.31665401063107, "grad_norm": 0.1045403853058815, "learning_rate": 1.9347298799347986e-05, "loss": 0.0204, "step": 115930 }, { "epoch": 2.3168538427720713, "grad_norm": 0.12239345163106918, "learning_rate": 1.9339135023311206e-05, "loss": 0.0285, "step": 115940 }, { "epoch": 2.317053674913073, "grad_norm": 0.11685465276241302, "learning_rate": 1.9330972557050257e-05, "loss": 0.0174, "step": 115950 }, { "epoch": 2.3172535070540747, "grad_norm": 0.14164716005325317, "learning_rate": 1.9322811400913837e-05, "loss": 0.0265, "step": 115960 }, { "epoch": 2.317453339195076, "grad_norm": 0.1152869164943695, "learning_rate": 1.9314651555250516e-05, "loss": 0.0233, "step": 115970 }, { "epoch": 2.3176531713360777, "grad_norm": 0.11201506108045578, "learning_rate": 1.9306493020408933e-05, "loss": 0.0325, "step": 115980 }, { "epoch": 2.3178530034770795, "grad_norm": 0.5240988731384277, "learning_rate": 1.9298335796737588e-05, "loss": 0.0197, "step": 115990 }, { "epoch": 2.3180528356180807, "grad_norm": 0.16302847862243652, "learning_rate": 1.9290179884584946e-05, "loss": 0.0267, "step": 116000 }, { "epoch": 2.3182526677590825, "grad_norm": 0.104206383228302, "learning_rate": 1.928202528429937e-05, "loss": 0.0286, "step": 116010 }, { "epoch": 2.3184524999000837, "grad_norm": 0.09770314395427704, "learning_rate": 1.9273871996229265e-05, "loss": 0.0307, "step": 116020 }, { "epoch": 2.3186523320410855, "grad_norm": 0.11307331919670105, "learning_rate": 1.9265720020722917e-05, "loss": 0.0201, "step": 116030 }, { "epoch": 2.318852164182087, "grad_norm": 0.1503528505563736, "learning_rate": 1.925756935812853e-05, "loss": 0.0228, "step": 116040 }, { "epoch": 2.3190519963230884, "grad_norm": 0.12841646373271942, "learning_rate": 1.9249420008794306e-05, "loss": 0.0295, "step": 116050 }, { "epoch": 2.31925182846409, "grad_norm": 0.10430365800857544, "learning_rate": 1.9241271973068365e-05, "loss": 0.0256, "step": 116060 }, { "epoch": 2.319451660605092, "grad_norm": 0.12808893620967865, "learning_rate": 1.9233125251298784e-05, "loss": 0.027, "step": 116070 }, { "epoch": 2.319651492746093, "grad_norm": 0.1225469559431076, "learning_rate": 1.9224979843833574e-05, "loss": 0.0262, "step": 116080 }, { "epoch": 2.319851324887095, "grad_norm": 0.12883391976356506, "learning_rate": 1.9216835751020683e-05, "loss": 0.0253, "step": 116090 }, { "epoch": 2.3200511570280966, "grad_norm": 0.0978679433465004, "learning_rate": 1.920869297320802e-05, "loss": 0.0257, "step": 116100 }, { "epoch": 2.320250989169098, "grad_norm": 0.2076440006494522, "learning_rate": 1.920055151074343e-05, "loss": 0.027, "step": 116110 }, { "epoch": 2.3204508213100996, "grad_norm": 0.11085154861211777, "learning_rate": 1.91924113639747e-05, "loss": 0.0215, "step": 116120 }, { "epoch": 2.320650653451101, "grad_norm": 0.20900516211986542, "learning_rate": 1.9184272533249577e-05, "loss": 0.0281, "step": 116130 }, { "epoch": 2.3208504855921026, "grad_norm": 0.13094136118888855, "learning_rate": 1.9176135018915703e-05, "loss": 0.028, "step": 116140 }, { "epoch": 2.3210503177331043, "grad_norm": 0.10588289797306061, "learning_rate": 1.9167998821320698e-05, "loss": 0.0202, "step": 116150 }, { "epoch": 2.3212501498741056, "grad_norm": 0.11672312766313553, "learning_rate": 1.9159863940812178e-05, "loss": 0.0309, "step": 116160 }, { "epoch": 2.3214499820151073, "grad_norm": 0.14488516747951508, "learning_rate": 1.9151730377737603e-05, "loss": 0.0299, "step": 116170 }, { "epoch": 2.321649814156109, "grad_norm": 0.2222309112548828, "learning_rate": 1.914359813244443e-05, "loss": 0.0212, "step": 116180 }, { "epoch": 2.3218496462971103, "grad_norm": 0.09095796942710876, "learning_rate": 1.913546720528006e-05, "loss": 0.0367, "step": 116190 }, { "epoch": 2.322049478438112, "grad_norm": 0.12634946405887604, "learning_rate": 1.912733759659183e-05, "loss": 0.0293, "step": 116200 }, { "epoch": 2.3222493105791138, "grad_norm": 0.12055807560682297, "learning_rate": 1.911920930672703e-05, "loss": 0.0251, "step": 116210 }, { "epoch": 2.322449142720115, "grad_norm": 0.11159668862819672, "learning_rate": 1.9111082336032876e-05, "loss": 0.0308, "step": 116220 }, { "epoch": 2.3226489748611168, "grad_norm": 0.1561397910118103, "learning_rate": 1.9102956684856556e-05, "loss": 0.025, "step": 116230 }, { "epoch": 2.322848807002118, "grad_norm": 0.1273602545261383, "learning_rate": 1.9094832353545128e-05, "loss": 0.0285, "step": 116240 }, { "epoch": 2.3230486391431198, "grad_norm": 0.09813088178634644, "learning_rate": 1.9086709342445713e-05, "loss": 0.0311, "step": 116250 }, { "epoch": 2.3232484712841215, "grad_norm": 0.11503671854734421, "learning_rate": 1.9078587651905293e-05, "loss": 0.025, "step": 116260 }, { "epoch": 2.3234483034251228, "grad_norm": 0.09897608309984207, "learning_rate": 1.907046728227079e-05, "loss": 0.0225, "step": 116270 }, { "epoch": 2.3236481355661245, "grad_norm": 0.11722198873758316, "learning_rate": 1.9062348233889094e-05, "loss": 0.0271, "step": 116280 }, { "epoch": 2.323847967707126, "grad_norm": 0.13906724750995636, "learning_rate": 1.905423050710708e-05, "loss": 0.0197, "step": 116290 }, { "epoch": 2.3240477998481275, "grad_norm": 0.1048894077539444, "learning_rate": 1.904611410227147e-05, "loss": 0.0228, "step": 116300 }, { "epoch": 2.324247631989129, "grad_norm": 0.11660993844270706, "learning_rate": 1.9037999019728997e-05, "loss": 0.0309, "step": 116310 }, { "epoch": 2.324447464130131, "grad_norm": 0.08889268338680267, "learning_rate": 1.902988525982633e-05, "loss": 0.0255, "step": 116320 }, { "epoch": 2.324647296271132, "grad_norm": 0.13528302311897278, "learning_rate": 1.9021772822910074e-05, "loss": 0.0259, "step": 116330 }, { "epoch": 2.324847128412134, "grad_norm": 0.12622353434562683, "learning_rate": 1.9013661709326773e-05, "loss": 0.0217, "step": 116340 }, { "epoch": 2.325046960553135, "grad_norm": 0.1237514317035675, "learning_rate": 1.900555191942292e-05, "loss": 0.0257, "step": 116350 }, { "epoch": 2.325246792694137, "grad_norm": 0.10769964009523392, "learning_rate": 1.899744345354495e-05, "loss": 0.022, "step": 116360 }, { "epoch": 2.3254466248351386, "grad_norm": 0.08943280577659607, "learning_rate": 1.898933631203924e-05, "loss": 0.0276, "step": 116370 }, { "epoch": 2.32564645697614, "grad_norm": 0.1306794434785843, "learning_rate": 1.8981230495252117e-05, "loss": 0.0271, "step": 116380 }, { "epoch": 2.3258462891171416, "grad_norm": 0.10404790192842484, "learning_rate": 1.897312600352984e-05, "loss": 0.0267, "step": 116390 }, { "epoch": 2.326046121258143, "grad_norm": 0.0905349850654602, "learning_rate": 1.8965022837218643e-05, "loss": 0.0202, "step": 116400 }, { "epoch": 2.3262459533991446, "grad_norm": 0.10839173942804337, "learning_rate": 1.895692099666464e-05, "loss": 0.0215, "step": 116410 }, { "epoch": 2.3264457855401464, "grad_norm": 0.10279950499534607, "learning_rate": 1.8948820482213924e-05, "loss": 0.0247, "step": 116420 }, { "epoch": 2.326645617681148, "grad_norm": 0.10896944254636765, "learning_rate": 1.8940721294212592e-05, "loss": 0.0275, "step": 116430 }, { "epoch": 2.3268454498221494, "grad_norm": 0.14165297150611877, "learning_rate": 1.8932623433006568e-05, "loss": 0.0248, "step": 116440 }, { "epoch": 2.327045281963151, "grad_norm": 0.11044670641422272, "learning_rate": 1.8924526898941802e-05, "loss": 0.0261, "step": 116450 }, { "epoch": 2.3272451141041524, "grad_norm": 0.10654391348361969, "learning_rate": 1.8916431692364157e-05, "loss": 0.0297, "step": 116460 }, { "epoch": 2.327444946245154, "grad_norm": 0.12765569984912872, "learning_rate": 1.8908337813619447e-05, "loss": 0.0314, "step": 116470 }, { "epoch": 2.327644778386156, "grad_norm": 0.11857704073190689, "learning_rate": 1.890024526305343e-05, "loss": 0.0297, "step": 116480 }, { "epoch": 2.327844610527157, "grad_norm": 0.09139220416545868, "learning_rate": 1.8892154041011807e-05, "loss": 0.03, "step": 116490 }, { "epoch": 2.328044442668159, "grad_norm": 0.10352444648742676, "learning_rate": 1.8884064147840225e-05, "loss": 0.0249, "step": 116500 }, { "epoch": 2.32824427480916, "grad_norm": 0.10322664678096771, "learning_rate": 1.887597558388423e-05, "loss": 0.026, "step": 116510 }, { "epoch": 2.328444106950162, "grad_norm": 0.10869339108467102, "learning_rate": 1.8867888349489406e-05, "loss": 0.0225, "step": 116520 }, { "epoch": 2.3286439390911635, "grad_norm": 0.15160860121250153, "learning_rate": 1.8859802445001208e-05, "loss": 0.0259, "step": 116530 }, { "epoch": 2.328843771232165, "grad_norm": 0.1168728917837143, "learning_rate": 1.885171787076503e-05, "loss": 0.0271, "step": 116540 }, { "epoch": 2.3290436033731665, "grad_norm": 0.12388799339532852, "learning_rate": 1.8843634627126223e-05, "loss": 0.0281, "step": 116550 }, { "epoch": 2.3292434355141682, "grad_norm": 0.11063773185014725, "learning_rate": 1.8835552714430148e-05, "loss": 0.0255, "step": 116560 }, { "epoch": 2.3294432676551695, "grad_norm": 0.16478894650936127, "learning_rate": 1.882747213302199e-05, "loss": 0.0286, "step": 116570 }, { "epoch": 2.3296430997961712, "grad_norm": 0.12287848442792892, "learning_rate": 1.8819392883246956e-05, "loss": 0.0316, "step": 116580 }, { "epoch": 2.329842931937173, "grad_norm": 0.09389607608318329, "learning_rate": 1.8811314965450178e-05, "loss": 0.0245, "step": 116590 }, { "epoch": 2.3300427640781742, "grad_norm": 0.11309286952018738, "learning_rate": 1.8803238379976727e-05, "loss": 0.025, "step": 116600 }, { "epoch": 2.330242596219176, "grad_norm": 0.12262942641973495, "learning_rate": 1.8795163127171627e-05, "loss": 0.0324, "step": 116610 }, { "epoch": 2.3304424283601772, "grad_norm": 0.12641490995883942, "learning_rate": 1.8787089207379834e-05, "loss": 0.0304, "step": 116620 }, { "epoch": 2.330642260501179, "grad_norm": 0.11901024729013443, "learning_rate": 1.8779016620946262e-05, "loss": 0.024, "step": 116630 }, { "epoch": 2.3308420926421807, "grad_norm": 0.11616619676351547, "learning_rate": 1.8770945368215715e-05, "loss": 0.0292, "step": 116640 }, { "epoch": 2.331041924783182, "grad_norm": 0.08887242525815964, "learning_rate": 1.8762875449533024e-05, "loss": 0.0258, "step": 116650 }, { "epoch": 2.3312417569241837, "grad_norm": 0.1662648469209671, "learning_rate": 1.875480686524293e-05, "loss": 0.0262, "step": 116660 }, { "epoch": 2.3314415890651854, "grad_norm": 0.13798663020133972, "learning_rate": 1.8746739615690073e-05, "loss": 0.0253, "step": 116670 }, { "epoch": 2.3316414212061867, "grad_norm": 0.15277661383152008, "learning_rate": 1.873867370121908e-05, "loss": 0.0263, "step": 116680 }, { "epoch": 2.3318412533471884, "grad_norm": 0.12137921899557114, "learning_rate": 1.873060912217452e-05, "loss": 0.0324, "step": 116690 }, { "epoch": 2.33204108548819, "grad_norm": 0.12768302857875824, "learning_rate": 1.87225458789009e-05, "loss": 0.0254, "step": 116700 }, { "epoch": 2.3322409176291914, "grad_norm": 0.12628990411758423, "learning_rate": 1.8714483971742654e-05, "loss": 0.0352, "step": 116710 }, { "epoch": 2.332440749770193, "grad_norm": 0.11227952688932419, "learning_rate": 1.8706423401044182e-05, "loss": 0.0251, "step": 116720 }, { "epoch": 2.3326405819111944, "grad_norm": 0.09735724329948425, "learning_rate": 1.869836416714981e-05, "loss": 0.0222, "step": 116730 }, { "epoch": 2.332840414052196, "grad_norm": 0.18621642887592316, "learning_rate": 1.8690306270403823e-05, "loss": 0.0232, "step": 116740 }, { "epoch": 2.333040246193198, "grad_norm": 0.15146265923976898, "learning_rate": 1.8682249711150434e-05, "loss": 0.0266, "step": 116750 }, { "epoch": 2.333240078334199, "grad_norm": 0.11061285436153412, "learning_rate": 1.8674194489733805e-05, "loss": 0.0224, "step": 116760 }, { "epoch": 2.333439910475201, "grad_norm": 0.15656650066375732, "learning_rate": 1.8666140606498057e-05, "loss": 0.0229, "step": 116770 }, { "epoch": 2.3336397426162026, "grad_norm": 0.12204601615667343, "learning_rate": 1.865808806178718e-05, "loss": 0.0228, "step": 116780 }, { "epoch": 2.333839574757204, "grad_norm": 0.11427733302116394, "learning_rate": 1.865003685594523e-05, "loss": 0.0204, "step": 116790 }, { "epoch": 2.3340394068982055, "grad_norm": 0.0971449539065361, "learning_rate": 1.864198698931613e-05, "loss": 0.0233, "step": 116800 }, { "epoch": 2.3342392390392073, "grad_norm": 0.11286963522434235, "learning_rate": 1.8633938462243716e-05, "loss": 0.0242, "step": 116810 }, { "epoch": 2.3344390711802085, "grad_norm": 0.13668115437030792, "learning_rate": 1.8625891275071818e-05, "loss": 0.0281, "step": 116820 }, { "epoch": 2.3346389033212103, "grad_norm": 0.1405164897441864, "learning_rate": 1.8617845428144238e-05, "loss": 0.0236, "step": 116830 }, { "epoch": 2.3348387354622115, "grad_norm": 0.13757233321666718, "learning_rate": 1.8609800921804638e-05, "loss": 0.0285, "step": 116840 }, { "epoch": 2.3350385676032133, "grad_norm": 0.13134993612766266, "learning_rate": 1.8601757756396676e-05, "loss": 0.0267, "step": 116850 }, { "epoch": 2.335238399744215, "grad_norm": 0.1257685124874115, "learning_rate": 1.859371593226394e-05, "loss": 0.0236, "step": 116860 }, { "epoch": 2.3354382318852163, "grad_norm": 0.12399864941835403, "learning_rate": 1.8585675449749967e-05, "loss": 0.0269, "step": 116870 }, { "epoch": 2.335638064026218, "grad_norm": 0.12594589591026306, "learning_rate": 1.8577636309198225e-05, "loss": 0.0256, "step": 116880 }, { "epoch": 2.3358378961672197, "grad_norm": 0.1227920725941658, "learning_rate": 1.856959851095214e-05, "loss": 0.0176, "step": 116890 }, { "epoch": 2.336037728308221, "grad_norm": 0.15089093148708344, "learning_rate": 1.856156205535508e-05, "loss": 0.0242, "step": 116900 }, { "epoch": 2.3362375604492227, "grad_norm": 0.10513366758823395, "learning_rate": 1.8553526942750305e-05, "loss": 0.0318, "step": 116910 }, { "epoch": 2.3364373925902244, "grad_norm": 0.11342473328113556, "learning_rate": 1.8545493173481097e-05, "loss": 0.0298, "step": 116920 }, { "epoch": 2.3366372247312257, "grad_norm": 0.14711390435695648, "learning_rate": 1.853746074789067e-05, "loss": 0.031, "step": 116930 }, { "epoch": 2.3368370568722274, "grad_norm": 0.12422911077737808, "learning_rate": 1.8529429666322092e-05, "loss": 0.02, "step": 116940 }, { "epoch": 2.3370368890132287, "grad_norm": 0.09951969981193542, "learning_rate": 1.8521399929118465e-05, "loss": 0.0275, "step": 116950 }, { "epoch": 2.3372367211542304, "grad_norm": 0.0949556976556778, "learning_rate": 1.8513371536622815e-05, "loss": 0.0267, "step": 116960 }, { "epoch": 2.337436553295232, "grad_norm": 0.08908507972955704, "learning_rate": 1.8505344489178082e-05, "loss": 0.0199, "step": 116970 }, { "epoch": 2.3376363854362334, "grad_norm": 0.13606597483158112, "learning_rate": 1.8497318787127178e-05, "loss": 0.0251, "step": 116980 }, { "epoch": 2.337836217577235, "grad_norm": 0.14303869009017944, "learning_rate": 1.8489294430812944e-05, "loss": 0.0204, "step": 116990 }, { "epoch": 2.338036049718237, "grad_norm": 0.10910051316022873, "learning_rate": 1.848127142057816e-05, "loss": 0.0234, "step": 117000 }, { "epoch": 2.338235881859238, "grad_norm": 0.12529873847961426, "learning_rate": 1.847324975676556e-05, "loss": 0.0262, "step": 117010 }, { "epoch": 2.33843571400024, "grad_norm": 0.13322801887989044, "learning_rate": 1.846522943971782e-05, "loss": 0.0299, "step": 117020 }, { "epoch": 2.3386355461412416, "grad_norm": 0.1396370381116867, "learning_rate": 1.8457210469777558e-05, "loss": 0.027, "step": 117030 }, { "epoch": 2.338835378282243, "grad_norm": 0.18176230788230896, "learning_rate": 1.8449192847287305e-05, "loss": 0.0222, "step": 117040 }, { "epoch": 2.3390352104232446, "grad_norm": 0.10886092483997345, "learning_rate": 1.8441176572589553e-05, "loss": 0.0246, "step": 117050 }, { "epoch": 2.339235042564246, "grad_norm": 0.1122555062174797, "learning_rate": 1.8433161646026796e-05, "loss": 0.0264, "step": 117060 }, { "epoch": 2.3394348747052476, "grad_norm": 0.13318558037281036, "learning_rate": 1.8425148067941372e-05, "loss": 0.031, "step": 117070 }, { "epoch": 2.3396347068462493, "grad_norm": 0.10701985657215118, "learning_rate": 1.8417135838675615e-05, "loss": 0.0224, "step": 117080 }, { "epoch": 2.3398345389872506, "grad_norm": 0.09157147258520126, "learning_rate": 1.8409124958571793e-05, "loss": 0.0184, "step": 117090 }, { "epoch": 2.3400343711282523, "grad_norm": 0.1098727434873581, "learning_rate": 1.840111542797212e-05, "loss": 0.0268, "step": 117100 }, { "epoch": 2.340234203269254, "grad_norm": 0.1261776238679886, "learning_rate": 1.839310724721875e-05, "loss": 0.0282, "step": 117110 }, { "epoch": 2.3404340354102553, "grad_norm": 0.14916794002056122, "learning_rate": 1.8385100416653766e-05, "loss": 0.0336, "step": 117120 }, { "epoch": 2.340633867551257, "grad_norm": 0.09496814012527466, "learning_rate": 1.837709493661924e-05, "loss": 0.025, "step": 117130 }, { "epoch": 2.3408336996922587, "grad_norm": 0.11463988572359085, "learning_rate": 1.8369090807457084e-05, "loss": 0.0198, "step": 117140 }, { "epoch": 2.34103353183326, "grad_norm": 0.11751764267683029, "learning_rate": 1.8361088029509283e-05, "loss": 0.0206, "step": 117150 }, { "epoch": 2.3412333639742617, "grad_norm": 0.10056204348802567, "learning_rate": 1.835308660311768e-05, "loss": 0.0316, "step": 117160 }, { "epoch": 2.341433196115263, "grad_norm": 0.11277298629283905, "learning_rate": 1.8345086528624094e-05, "loss": 0.0266, "step": 117170 }, { "epoch": 2.3416330282562647, "grad_norm": 0.1012430265545845, "learning_rate": 1.8337087806370225e-05, "loss": 0.0216, "step": 117180 }, { "epoch": 2.3418328603972665, "grad_norm": 0.11080954968929291, "learning_rate": 1.8329090436697817e-05, "loss": 0.0174, "step": 117190 }, { "epoch": 2.3420326925382677, "grad_norm": 0.1128249540925026, "learning_rate": 1.8321094419948496e-05, "loss": 0.0325, "step": 117200 }, { "epoch": 2.3422325246792695, "grad_norm": 0.14235535264015198, "learning_rate": 1.8313099756463814e-05, "loss": 0.0286, "step": 117210 }, { "epoch": 2.3424323568202707, "grad_norm": 0.14264936745166779, "learning_rate": 1.8305106446585286e-05, "loss": 0.0259, "step": 117220 }, { "epoch": 2.3426321889612725, "grad_norm": 0.13060003519058228, "learning_rate": 1.829711449065441e-05, "loss": 0.0215, "step": 117230 }, { "epoch": 2.342832021102274, "grad_norm": 0.13654258847236633, "learning_rate": 1.8289123889012554e-05, "loss": 0.022, "step": 117240 }, { "epoch": 2.343031853243276, "grad_norm": 0.1425987184047699, "learning_rate": 1.828113464200107e-05, "loss": 0.0253, "step": 117250 }, { "epoch": 2.343231685384277, "grad_norm": 0.10559332370758057, "learning_rate": 1.827314674996124e-05, "loss": 0.0314, "step": 117260 }, { "epoch": 2.343431517525279, "grad_norm": 0.10705661028623581, "learning_rate": 1.82651602132343e-05, "loss": 0.0222, "step": 117270 }, { "epoch": 2.34363134966628, "grad_norm": 0.13948753476142883, "learning_rate": 1.825717503216141e-05, "loss": 0.0286, "step": 117280 }, { "epoch": 2.343831181807282, "grad_norm": 0.11970193684101105, "learning_rate": 1.82491912070837e-05, "loss": 0.023, "step": 117290 }, { "epoch": 2.3440310139482836, "grad_norm": 0.11570905148983002, "learning_rate": 1.8241208738342224e-05, "loss": 0.0254, "step": 117300 }, { "epoch": 2.344230846089285, "grad_norm": 0.19479776918888092, "learning_rate": 1.8233227626277937e-05, "loss": 0.0219, "step": 117310 }, { "epoch": 2.3444306782302866, "grad_norm": 0.09828583896160126, "learning_rate": 1.822524787123182e-05, "loss": 0.0209, "step": 117320 }, { "epoch": 2.344630510371288, "grad_norm": 0.12066032737493515, "learning_rate": 1.821726947354476e-05, "loss": 0.0227, "step": 117330 }, { "epoch": 2.3448303425122896, "grad_norm": 0.13417769968509674, "learning_rate": 1.820929243355755e-05, "loss": 0.0323, "step": 117340 }, { "epoch": 2.3450301746532913, "grad_norm": 0.1323033720254898, "learning_rate": 1.8201316751610964e-05, "loss": 0.0244, "step": 117350 }, { "epoch": 2.3452300067942926, "grad_norm": 0.11522047221660614, "learning_rate": 1.8193342428045714e-05, "loss": 0.0269, "step": 117360 }, { "epoch": 2.3454298389352943, "grad_norm": 0.09989624470472336, "learning_rate": 1.8185369463202433e-05, "loss": 0.033, "step": 117370 }, { "epoch": 2.345629671076296, "grad_norm": 0.1403040736913681, "learning_rate": 1.8177397857421737e-05, "loss": 0.0299, "step": 117380 }, { "epoch": 2.3458295032172973, "grad_norm": 0.1886960119009018, "learning_rate": 1.8169427611044133e-05, "loss": 0.0285, "step": 117390 }, { "epoch": 2.346029335358299, "grad_norm": 0.13455916941165924, "learning_rate": 1.8161458724410112e-05, "loss": 0.0252, "step": 117400 }, { "epoch": 2.3462291674993008, "grad_norm": 0.11307799071073532, "learning_rate": 1.8153491197860083e-05, "loss": 0.031, "step": 117410 }, { "epoch": 2.346428999640302, "grad_norm": 0.12295113503932953, "learning_rate": 1.814552503173441e-05, "loss": 0.0231, "step": 117420 }, { "epoch": 2.3466288317813038, "grad_norm": 0.13382799923419952, "learning_rate": 1.8137560226373402e-05, "loss": 0.0222, "step": 117430 }, { "epoch": 2.346828663922305, "grad_norm": 0.12260399758815765, "learning_rate": 1.8129596782117276e-05, "loss": 0.0247, "step": 117440 }, { "epoch": 2.3470284960633068, "grad_norm": 0.13582298159599304, "learning_rate": 1.8121634699306206e-05, "loss": 0.021, "step": 117450 }, { "epoch": 2.3472283282043085, "grad_norm": 0.1020776629447937, "learning_rate": 1.8113673978280372e-05, "loss": 0.0247, "step": 117460 }, { "epoch": 2.3474281603453098, "grad_norm": 0.10526945441961288, "learning_rate": 1.8105714619379792e-05, "loss": 0.0209, "step": 117470 }, { "epoch": 2.3476279924863115, "grad_norm": 0.1734115034341812, "learning_rate": 1.8097756622944494e-05, "loss": 0.0312, "step": 117480 }, { "epoch": 2.347827824627313, "grad_norm": 0.13245005905628204, "learning_rate": 1.8089799989314433e-05, "loss": 0.0228, "step": 117490 }, { "epoch": 2.3480276567683145, "grad_norm": 0.27232593297958374, "learning_rate": 1.8081844718829487e-05, "loss": 0.0294, "step": 117500 }, { "epoch": 2.348227488909316, "grad_norm": 0.10599356144666672, "learning_rate": 1.80738908118295e-05, "loss": 0.0158, "step": 117510 }, { "epoch": 2.348427321050318, "grad_norm": 0.10917311906814575, "learning_rate": 1.8065938268654246e-05, "loss": 0.0258, "step": 117520 }, { "epoch": 2.348627153191319, "grad_norm": 0.1286390870809555, "learning_rate": 1.805798708964347e-05, "loss": 0.0191, "step": 117530 }, { "epoch": 2.348826985332321, "grad_norm": 0.12532728910446167, "learning_rate": 1.8050037275136765e-05, "loss": 0.0224, "step": 117540 }, { "epoch": 2.349026817473322, "grad_norm": 0.11410074681043625, "learning_rate": 1.8042088825473792e-05, "loss": 0.0359, "step": 117550 }, { "epoch": 2.349226649614324, "grad_norm": 0.20609882473945618, "learning_rate": 1.8034141740994075e-05, "loss": 0.0267, "step": 117560 }, { "epoch": 2.3494264817553256, "grad_norm": 0.0986391082406044, "learning_rate": 1.8026196022037117e-05, "loss": 0.0223, "step": 117570 }, { "epoch": 2.349626313896327, "grad_norm": 0.10940445214509964, "learning_rate": 1.8018251668942294e-05, "loss": 0.0284, "step": 117580 }, { "epoch": 2.3498261460373286, "grad_norm": 0.11204344034194946, "learning_rate": 1.8010308682049026e-05, "loss": 0.0229, "step": 117590 }, { "epoch": 2.3500259781783304, "grad_norm": 0.12383623421192169, "learning_rate": 1.8002367061696618e-05, "loss": 0.025, "step": 117600 }, { "epoch": 2.3502258103193316, "grad_norm": 0.13340066373348236, "learning_rate": 1.799442680822429e-05, "loss": 0.02, "step": 117610 }, { "epoch": 2.3504256424603334, "grad_norm": 0.0927710309624672, "learning_rate": 1.7986487921971256e-05, "loss": 0.0236, "step": 117620 }, { "epoch": 2.350625474601335, "grad_norm": 0.1327207386493683, "learning_rate": 1.7978550403276646e-05, "loss": 0.0306, "step": 117630 }, { "epoch": 2.3508253067423364, "grad_norm": 0.16439788043498993, "learning_rate": 1.7970614252479534e-05, "loss": 0.025, "step": 117640 }, { "epoch": 2.351025138883338, "grad_norm": 0.09947662055492401, "learning_rate": 1.7962679469918947e-05, "loss": 0.033, "step": 117650 }, { "epoch": 2.3512249710243394, "grad_norm": 0.15929101407527924, "learning_rate": 1.7954746055933836e-05, "loss": 0.0234, "step": 117660 }, { "epoch": 2.351424803165341, "grad_norm": 0.18323194980621338, "learning_rate": 1.7946814010863107e-05, "loss": 0.0299, "step": 117670 }, { "epoch": 2.351624635306343, "grad_norm": 0.1238221749663353, "learning_rate": 1.7938883335045592e-05, "loss": 0.0242, "step": 117680 }, { "epoch": 2.351824467447344, "grad_norm": 0.14577685296535492, "learning_rate": 1.7930954028820092e-05, "loss": 0.0229, "step": 117690 }, { "epoch": 2.352024299588346, "grad_norm": 0.09871530532836914, "learning_rate": 1.7923026092525334e-05, "loss": 0.0308, "step": 117700 }, { "epoch": 2.3522241317293475, "grad_norm": 0.12203630059957504, "learning_rate": 1.7915099526499964e-05, "loss": 0.0286, "step": 117710 }, { "epoch": 2.352423963870349, "grad_norm": 0.1299629807472229, "learning_rate": 1.7907174331082576e-05, "loss": 0.0465, "step": 117720 }, { "epoch": 2.3526237960113505, "grad_norm": 0.10622797906398773, "learning_rate": 1.789925050661178e-05, "loss": 0.0218, "step": 117730 }, { "epoch": 2.3528236281523522, "grad_norm": 0.13587675988674164, "learning_rate": 1.7891328053426016e-05, "loss": 0.0214, "step": 117740 }, { "epoch": 2.3530234602933535, "grad_norm": 0.11814180016517639, "learning_rate": 1.7883406971863732e-05, "loss": 0.0202, "step": 117750 }, { "epoch": 2.3532232924343552, "grad_norm": 0.10299959033727646, "learning_rate": 1.78754872622633e-05, "loss": 0.0175, "step": 117760 }, { "epoch": 2.3534231245753565, "grad_norm": 0.15793873369693756, "learning_rate": 1.7867568924963042e-05, "loss": 0.0273, "step": 117770 }, { "epoch": 2.3536229567163582, "grad_norm": 0.12645979225635529, "learning_rate": 1.7859651960301205e-05, "loss": 0.0238, "step": 117780 }, { "epoch": 2.35382278885736, "grad_norm": 0.10789724439382553, "learning_rate": 1.7851736368615995e-05, "loss": 0.0209, "step": 117790 }, { "epoch": 2.3540226209983612, "grad_norm": 0.1190505400300026, "learning_rate": 1.7843822150245566e-05, "loss": 0.0239, "step": 117800 }, { "epoch": 2.354222453139363, "grad_norm": 0.18281276524066925, "learning_rate": 1.783590930552795e-05, "loss": 0.0235, "step": 117810 }, { "epoch": 2.3544222852803647, "grad_norm": 0.14329677820205688, "learning_rate": 1.7827997834801213e-05, "loss": 0.0245, "step": 117820 }, { "epoch": 2.354622117421366, "grad_norm": 0.11202836781740189, "learning_rate": 1.782008773840333e-05, "loss": 0.0213, "step": 117830 }, { "epoch": 2.3548219495623677, "grad_norm": 0.1285216510295868, "learning_rate": 1.7812179016672158e-05, "loss": 0.0405, "step": 117840 }, { "epoch": 2.3550217817033694, "grad_norm": 0.13590103387832642, "learning_rate": 1.7804271669945553e-05, "loss": 0.0243, "step": 117850 }, { "epoch": 2.3552216138443707, "grad_norm": 0.08585847169160843, "learning_rate": 1.7796365698561355e-05, "loss": 0.0189, "step": 117860 }, { "epoch": 2.3554214459853724, "grad_norm": 0.11833442747592926, "learning_rate": 1.7788461102857234e-05, "loss": 0.0319, "step": 117870 }, { "epoch": 2.3556212781263737, "grad_norm": 0.10816412419080734, "learning_rate": 1.7780557883170886e-05, "loss": 0.0189, "step": 117880 }, { "epoch": 2.3558211102673754, "grad_norm": 0.11599405109882355, "learning_rate": 1.7772656039839906e-05, "loss": 0.0302, "step": 117890 }, { "epoch": 2.356020942408377, "grad_norm": 0.09842474013566971, "learning_rate": 1.7764755573201863e-05, "loss": 0.0258, "step": 117900 }, { "epoch": 2.3562207745493784, "grad_norm": 0.147963285446167, "learning_rate": 1.7756856483594243e-05, "loss": 0.0205, "step": 117910 }, { "epoch": 2.35642060669038, "grad_norm": 0.10577448457479477, "learning_rate": 1.7748958771354478e-05, "loss": 0.0236, "step": 117920 }, { "epoch": 2.3566204388313814, "grad_norm": 0.11267861723899841, "learning_rate": 1.774106243681997e-05, "loss": 0.0264, "step": 117930 }, { "epoch": 2.356820270972383, "grad_norm": 0.13458722829818726, "learning_rate": 1.7733167480327978e-05, "loss": 0.0256, "step": 117940 }, { "epoch": 2.357020103113385, "grad_norm": 0.13460728526115417, "learning_rate": 1.772527390221581e-05, "loss": 0.0238, "step": 117950 }, { "epoch": 2.3572199352543866, "grad_norm": 0.14488916099071503, "learning_rate": 1.771738170282065e-05, "loss": 0.03, "step": 117960 }, { "epoch": 2.357419767395388, "grad_norm": 0.14135746657848358, "learning_rate": 1.7709490882479656e-05, "loss": 0.0266, "step": 117970 }, { "epoch": 2.3576195995363896, "grad_norm": 0.08515457063913345, "learning_rate": 1.7701601441529876e-05, "loss": 0.0231, "step": 117980 }, { "epoch": 2.357819431677391, "grad_norm": 0.12075550109148026, "learning_rate": 1.769371338030833e-05, "loss": 0.0241, "step": 117990 }, { "epoch": 2.3580192638183926, "grad_norm": 0.13248899579048157, "learning_rate": 1.7685826699152037e-05, "loss": 0.0249, "step": 118000 }, { "epoch": 2.3582190959593943, "grad_norm": 0.1105482205748558, "learning_rate": 1.7677941398397845e-05, "loss": 0.0231, "step": 118010 }, { "epoch": 2.3584189281003956, "grad_norm": 0.15218274295330048, "learning_rate": 1.7670057478382617e-05, "loss": 0.019, "step": 118020 }, { "epoch": 2.3586187602413973, "grad_norm": 0.15202103555202484, "learning_rate": 1.7662174939443145e-05, "loss": 0.0358, "step": 118030 }, { "epoch": 2.3588185923823985, "grad_norm": 0.1347905397415161, "learning_rate": 1.7654293781916153e-05, "loss": 0.0238, "step": 118040 }, { "epoch": 2.3590184245234003, "grad_norm": 0.13950853049755096, "learning_rate": 1.7646414006138308e-05, "loss": 0.0289, "step": 118050 }, { "epoch": 2.359218256664402, "grad_norm": 0.11796902865171432, "learning_rate": 1.7638535612446224e-05, "loss": 0.0315, "step": 118060 }, { "epoch": 2.3594180888054033, "grad_norm": 0.13728612661361694, "learning_rate": 1.763065860117646e-05, "loss": 0.0238, "step": 118070 }, { "epoch": 2.359617920946405, "grad_norm": 0.14250878989696503, "learning_rate": 1.7622782972665464e-05, "loss": 0.0235, "step": 118080 }, { "epoch": 2.3598177530874067, "grad_norm": 0.08743687719106674, "learning_rate": 1.7614908727249717e-05, "loss": 0.026, "step": 118090 }, { "epoch": 2.360017585228408, "grad_norm": 0.10014202445745468, "learning_rate": 1.7607035865265597e-05, "loss": 0.0259, "step": 118100 }, { "epoch": 2.3602174173694097, "grad_norm": 0.1473703235387802, "learning_rate": 1.7599164387049378e-05, "loss": 0.0228, "step": 118110 }, { "epoch": 2.3604172495104114, "grad_norm": 0.11782868951559067, "learning_rate": 1.7591294292937314e-05, "loss": 0.0216, "step": 118120 }, { "epoch": 2.3606170816514127, "grad_norm": 0.09751994907855988, "learning_rate": 1.758342558326566e-05, "loss": 0.03, "step": 118130 }, { "epoch": 2.3608169137924144, "grad_norm": 0.13778968155384064, "learning_rate": 1.75755582583705e-05, "loss": 0.0223, "step": 118140 }, { "epoch": 2.3610167459334157, "grad_norm": 0.19532328844070435, "learning_rate": 1.756769231858792e-05, "loss": 0.0241, "step": 118150 }, { "epoch": 2.3612165780744174, "grad_norm": 0.11864808946847916, "learning_rate": 1.7559827764253943e-05, "loss": 0.0313, "step": 118160 }, { "epoch": 2.361416410215419, "grad_norm": 0.12690913677215576, "learning_rate": 1.755196459570453e-05, "loss": 0.0258, "step": 118170 }, { "epoch": 2.3616162423564204, "grad_norm": 0.14536869525909424, "learning_rate": 1.754410281327558e-05, "loss": 0.0196, "step": 118180 }, { "epoch": 2.361816074497422, "grad_norm": 0.12098322808742523, "learning_rate": 1.753624241730294e-05, "loss": 0.0212, "step": 118190 }, { "epoch": 2.362015906638424, "grad_norm": 0.14209753274917603, "learning_rate": 1.7528383408122402e-05, "loss": 0.0274, "step": 118200 }, { "epoch": 2.362215738779425, "grad_norm": 0.10788597166538239, "learning_rate": 1.752052578606963e-05, "loss": 0.0209, "step": 118210 }, { "epoch": 2.362415570920427, "grad_norm": 0.1488129049539566, "learning_rate": 1.7512669551480354e-05, "loss": 0.0319, "step": 118220 }, { "epoch": 2.3626154030614286, "grad_norm": 0.13710732758045197, "learning_rate": 1.750481470469017e-05, "loss": 0.0228, "step": 118230 }, { "epoch": 2.36281523520243, "grad_norm": 0.1212160587310791, "learning_rate": 1.749696124603459e-05, "loss": 0.0256, "step": 118240 }, { "epoch": 2.3630150673434316, "grad_norm": 0.1136460155248642, "learning_rate": 1.748910917584911e-05, "loss": 0.0233, "step": 118250 }, { "epoch": 2.363214899484433, "grad_norm": 0.14843985438346863, "learning_rate": 1.7481258494469168e-05, "loss": 0.0289, "step": 118260 }, { "epoch": 2.3634147316254346, "grad_norm": 0.11221875995397568, "learning_rate": 1.7473409202230125e-05, "loss": 0.0255, "step": 118270 }, { "epoch": 2.3636145637664363, "grad_norm": 0.13707378506660461, "learning_rate": 1.746556129946728e-05, "loss": 0.0188, "step": 118280 }, { "epoch": 2.3638143959074376, "grad_norm": 0.09351257979869843, "learning_rate": 1.7457714786515895e-05, "loss": 0.036, "step": 118290 }, { "epoch": 2.3640142280484393, "grad_norm": 0.10358802229166031, "learning_rate": 1.744986966371115e-05, "loss": 0.0288, "step": 118300 }, { "epoch": 2.364214060189441, "grad_norm": 0.1302049160003662, "learning_rate": 1.7442025931388178e-05, "loss": 0.031, "step": 118310 }, { "epoch": 2.3644138923304423, "grad_norm": 0.13573025166988373, "learning_rate": 1.743418358988205e-05, "loss": 0.0243, "step": 118320 }, { "epoch": 2.364613724471444, "grad_norm": 0.08905129879713058, "learning_rate": 1.7426342639527783e-05, "loss": 0.0314, "step": 118330 }, { "epoch": 2.3648135566124457, "grad_norm": 0.16102585196495056, "learning_rate": 1.7418503080660285e-05, "loss": 0.0258, "step": 118340 }, { "epoch": 2.365013388753447, "grad_norm": 0.0997769832611084, "learning_rate": 1.7410664913614505e-05, "loss": 0.0175, "step": 118350 }, { "epoch": 2.3652132208944487, "grad_norm": 0.11965082585811615, "learning_rate": 1.7402828138725246e-05, "loss": 0.0199, "step": 118360 }, { "epoch": 2.36541305303545, "grad_norm": 0.15178871154785156, "learning_rate": 1.7394992756327306e-05, "loss": 0.0234, "step": 118370 }, { "epoch": 2.3656128851764517, "grad_norm": 0.13456082344055176, "learning_rate": 1.7387158766755363e-05, "loss": 0.0308, "step": 118380 }, { "epoch": 2.3658127173174535, "grad_norm": 0.13838452100753784, "learning_rate": 1.737932617034406e-05, "loss": 0.0249, "step": 118390 }, { "epoch": 2.3660125494584547, "grad_norm": 0.131048783659935, "learning_rate": 1.737149496742806e-05, "loss": 0.0256, "step": 118400 }, { "epoch": 2.3662123815994565, "grad_norm": 0.12442990392446518, "learning_rate": 1.7363665158341834e-05, "loss": 0.0284, "step": 118410 }, { "epoch": 2.366412213740458, "grad_norm": 0.10150083154439926, "learning_rate": 1.735583674341988e-05, "loss": 0.0231, "step": 118420 }, { "epoch": 2.3666120458814595, "grad_norm": 0.19947652518749237, "learning_rate": 1.7348009722996612e-05, "loss": 0.0249, "step": 118430 }, { "epoch": 2.366811878022461, "grad_norm": 0.0843118205666542, "learning_rate": 1.734018409740638e-05, "loss": 0.0255, "step": 118440 }, { "epoch": 2.367011710163463, "grad_norm": 0.11822833865880966, "learning_rate": 1.733235986698349e-05, "loss": 0.0234, "step": 118450 }, { "epoch": 2.367211542304464, "grad_norm": 0.16640980541706085, "learning_rate": 1.7324537032062167e-05, "loss": 0.0212, "step": 118460 }, { "epoch": 2.367411374445466, "grad_norm": 0.13089723885059357, "learning_rate": 1.7316715592976617e-05, "loss": 0.0306, "step": 118470 }, { "epoch": 2.367611206586467, "grad_norm": 0.14718548953533173, "learning_rate": 1.7308895550060905e-05, "loss": 0.0265, "step": 118480 }, { "epoch": 2.367811038727469, "grad_norm": 0.10229359567165375, "learning_rate": 1.730107690364913e-05, "loss": 0.0231, "step": 118490 }, { "epoch": 2.3680108708684706, "grad_norm": 0.09708797186613083, "learning_rate": 1.7293259654075302e-05, "loss": 0.0208, "step": 118500 }, { "epoch": 2.368210703009472, "grad_norm": 0.12581098079681396, "learning_rate": 1.7285443801673316e-05, "loss": 0.032, "step": 118510 }, { "epoch": 2.3684105351504736, "grad_norm": 0.09055133908987045, "learning_rate": 1.727762934677706e-05, "loss": 0.0306, "step": 118520 }, { "epoch": 2.3686103672914753, "grad_norm": 0.1176796406507492, "learning_rate": 1.72698162897204e-05, "loss": 0.0276, "step": 118530 }, { "epoch": 2.3688101994324766, "grad_norm": 0.09882427006959915, "learning_rate": 1.7262004630837036e-05, "loss": 0.0271, "step": 118540 }, { "epoch": 2.3690100315734783, "grad_norm": 0.12047922611236572, "learning_rate": 1.7254194370460697e-05, "loss": 0.0281, "step": 118550 }, { "epoch": 2.36920986371448, "grad_norm": 0.144328311085701, "learning_rate": 1.7246385508925022e-05, "loss": 0.0304, "step": 118560 }, { "epoch": 2.3694096958554813, "grad_norm": 0.20814305543899536, "learning_rate": 1.7238578046563588e-05, "loss": 0.0216, "step": 118570 }, { "epoch": 2.369609527996483, "grad_norm": 0.12251286953687668, "learning_rate": 1.7230771983709908e-05, "loss": 0.0186, "step": 118580 }, { "epoch": 2.3698093601374843, "grad_norm": 0.14771293103694916, "learning_rate": 1.7222967320697452e-05, "loss": 0.0253, "step": 118590 }, { "epoch": 2.370009192278486, "grad_norm": 0.12959423661231995, "learning_rate": 1.721516405785964e-05, "loss": 0.0334, "step": 118600 }, { "epoch": 2.3702090244194878, "grad_norm": 0.10428027808666229, "learning_rate": 1.7207362195529753e-05, "loss": 0.027, "step": 118610 }, { "epoch": 2.370408856560489, "grad_norm": 0.11107977479696274, "learning_rate": 1.719956173404113e-05, "loss": 0.0224, "step": 118620 }, { "epoch": 2.3706086887014908, "grad_norm": 0.09879035502672195, "learning_rate": 1.719176267372699e-05, "loss": 0.0234, "step": 118630 }, { "epoch": 2.3708085208424925, "grad_norm": 0.11136005073785782, "learning_rate": 1.7183965014920463e-05, "loss": 0.0248, "step": 118640 }, { "epoch": 2.3710083529834938, "grad_norm": 0.09361228346824646, "learning_rate": 1.7176168757954662e-05, "loss": 0.0244, "step": 118650 }, { "epoch": 2.3712081851244955, "grad_norm": 0.13069140911102295, "learning_rate": 1.716837390316264e-05, "loss": 0.0251, "step": 118660 }, { "epoch": 2.371408017265497, "grad_norm": 0.129459410905838, "learning_rate": 1.716058045087736e-05, "loss": 0.0217, "step": 118670 }, { "epoch": 2.3716078494064985, "grad_norm": 0.11490751057863235, "learning_rate": 1.7152788401431763e-05, "loss": 0.0203, "step": 118680 }, { "epoch": 2.3718076815475, "grad_norm": 0.10972586274147034, "learning_rate": 1.7144997755158703e-05, "loss": 0.0187, "step": 118690 }, { "epoch": 2.3720075136885015, "grad_norm": 0.13634096086025238, "learning_rate": 1.713720851239098e-05, "loss": 0.0265, "step": 118700 }, { "epoch": 2.372207345829503, "grad_norm": 0.19704899191856384, "learning_rate": 1.7129420673461338e-05, "loss": 0.0276, "step": 118710 }, { "epoch": 2.372407177970505, "grad_norm": 0.12090735137462616, "learning_rate": 1.7121634238702462e-05, "loss": 0.0283, "step": 118720 }, { "epoch": 2.372607010111506, "grad_norm": 0.13945016264915466, "learning_rate": 1.7113849208446968e-05, "loss": 0.0257, "step": 118730 }, { "epoch": 2.372806842252508, "grad_norm": 0.11303146183490753, "learning_rate": 1.710606558302744e-05, "loss": 0.0259, "step": 118740 }, { "epoch": 2.373006674393509, "grad_norm": 0.07687316834926605, "learning_rate": 1.7098283362776336e-05, "loss": 0.0226, "step": 118750 }, { "epoch": 2.373206506534511, "grad_norm": 0.1241925060749054, "learning_rate": 1.7090502548026133e-05, "loss": 0.0196, "step": 118760 }, { "epoch": 2.3734063386755127, "grad_norm": 0.131161168217659, "learning_rate": 1.7082723139109225e-05, "loss": 0.0289, "step": 118770 }, { "epoch": 2.3736061708165144, "grad_norm": 0.1251497119665146, "learning_rate": 1.7074945136357896e-05, "loss": 0.0238, "step": 118780 }, { "epoch": 2.3738060029575156, "grad_norm": 0.17914769053459167, "learning_rate": 1.7067168540104407e-05, "loss": 0.0255, "step": 118790 }, { "epoch": 2.3740058350985174, "grad_norm": 0.09846366196870804, "learning_rate": 1.7059393350681014e-05, "loss": 0.024, "step": 118800 }, { "epoch": 2.3742056672395186, "grad_norm": 0.1285005360841751, "learning_rate": 1.7051619568419802e-05, "loss": 0.0264, "step": 118810 }, { "epoch": 2.3744054993805204, "grad_norm": 0.13107478618621826, "learning_rate": 1.704384719365288e-05, "loss": 0.0283, "step": 118820 }, { "epoch": 2.374605331521522, "grad_norm": 0.13151633739471436, "learning_rate": 1.703607622671226e-05, "loss": 0.0205, "step": 118830 }, { "epoch": 2.3748051636625234, "grad_norm": 0.10619368404150009, "learning_rate": 1.7028306667929905e-05, "loss": 0.0232, "step": 118840 }, { "epoch": 2.375004995803525, "grad_norm": 0.11759471893310547, "learning_rate": 1.7020538517637726e-05, "loss": 0.0149, "step": 118850 }, { "epoch": 2.3752048279445264, "grad_norm": 0.16867172718048096, "learning_rate": 1.7012771776167547e-05, "loss": 0.0288, "step": 118860 }, { "epoch": 2.375404660085528, "grad_norm": 0.09994865208864212, "learning_rate": 1.700500644385118e-05, "loss": 0.0266, "step": 118870 }, { "epoch": 2.37560449222653, "grad_norm": 0.07820593565702438, "learning_rate": 1.699724252102029e-05, "loss": 0.0237, "step": 118880 }, { "epoch": 2.375804324367531, "grad_norm": 0.12917488813400269, "learning_rate": 1.6989480008006585e-05, "loss": 0.0245, "step": 118890 }, { "epoch": 2.376004156508533, "grad_norm": 0.16535045206546783, "learning_rate": 1.6981718905141674e-05, "loss": 0.0243, "step": 118900 }, { "epoch": 2.3762039886495345, "grad_norm": 0.09171412885189056, "learning_rate": 1.6973959212757056e-05, "loss": 0.028, "step": 118910 }, { "epoch": 2.376403820790536, "grad_norm": 0.14818647503852844, "learning_rate": 1.6966200931184228e-05, "loss": 0.0235, "step": 118920 }, { "epoch": 2.3766036529315375, "grad_norm": 0.08521809428930283, "learning_rate": 1.6958444060754614e-05, "loss": 0.0259, "step": 118930 }, { "epoch": 2.3768034850725392, "grad_norm": 0.11235898733139038, "learning_rate": 1.695068860179957e-05, "loss": 0.0183, "step": 118940 }, { "epoch": 2.3770033172135405, "grad_norm": 0.1582939326763153, "learning_rate": 1.694293455465039e-05, "loss": 0.0191, "step": 118950 }, { "epoch": 2.3772031493545422, "grad_norm": 0.09828189760446548, "learning_rate": 1.6935181919638323e-05, "loss": 0.0299, "step": 118960 }, { "epoch": 2.3774029814955435, "grad_norm": 0.09850288927555084, "learning_rate": 1.692743069709454e-05, "loss": 0.0242, "step": 118970 }, { "epoch": 2.3776028136365452, "grad_norm": 0.16400614380836487, "learning_rate": 1.6919680887350166e-05, "loss": 0.0309, "step": 118980 }, { "epoch": 2.377802645777547, "grad_norm": 0.11983848363161087, "learning_rate": 1.691193249073625e-05, "loss": 0.0247, "step": 118990 }, { "epoch": 2.3780024779185482, "grad_norm": 0.14824508130550385, "learning_rate": 1.6904185507583815e-05, "loss": 0.0302, "step": 119000 }, { "epoch": 2.37820231005955, "grad_norm": 0.11118777841329575, "learning_rate": 1.6896439938223752e-05, "loss": 0.0226, "step": 119010 }, { "epoch": 2.3784021422005517, "grad_norm": 0.15535515546798706, "learning_rate": 1.6888695782986946e-05, "loss": 0.0305, "step": 119020 }, { "epoch": 2.378601974341553, "grad_norm": 0.09847230464220047, "learning_rate": 1.6880953042204268e-05, "loss": 0.026, "step": 119030 }, { "epoch": 2.3788018064825547, "grad_norm": 0.09908954054117203, "learning_rate": 1.687321171620641e-05, "loss": 0.024, "step": 119040 }, { "epoch": 2.3790016386235564, "grad_norm": 0.12207195907831192, "learning_rate": 1.68654718053241e-05, "loss": 0.0275, "step": 119050 }, { "epoch": 2.3792014707645577, "grad_norm": 0.10697299987077713, "learning_rate": 1.685773330988796e-05, "loss": 0.0323, "step": 119060 }, { "epoch": 2.3794013029055594, "grad_norm": 0.1453356146812439, "learning_rate": 1.684999623022857e-05, "loss": 0.0222, "step": 119070 }, { "epoch": 2.3796011350465607, "grad_norm": 0.10129453241825104, "learning_rate": 1.6842260566676436e-05, "loss": 0.0234, "step": 119080 }, { "epoch": 2.3798009671875624, "grad_norm": 0.19209115207195282, "learning_rate": 1.6834526319562028e-05, "loss": 0.0291, "step": 119090 }, { "epoch": 2.380000799328564, "grad_norm": 0.15131714940071106, "learning_rate": 1.682679348921574e-05, "loss": 0.0226, "step": 119100 }, { "epoch": 2.3802006314695654, "grad_norm": 0.1083296611905098, "learning_rate": 1.6819062075967863e-05, "loss": 0.0374, "step": 119110 }, { "epoch": 2.380400463610567, "grad_norm": 0.1456626057624817, "learning_rate": 1.681133208014871e-05, "loss": 0.0213, "step": 119120 }, { "epoch": 2.380600295751569, "grad_norm": 0.09796067327260971, "learning_rate": 1.6803603502088484e-05, "loss": 0.0259, "step": 119130 }, { "epoch": 2.38080012789257, "grad_norm": 0.13565167784690857, "learning_rate": 1.679587634211735e-05, "loss": 0.0334, "step": 119140 }, { "epoch": 2.380999960033572, "grad_norm": 0.11562889069318771, "learning_rate": 1.6788150600565343e-05, "loss": 0.0204, "step": 119150 }, { "epoch": 2.3811997921745736, "grad_norm": 0.15139240026474, "learning_rate": 1.6780426277762558e-05, "loss": 0.0244, "step": 119160 }, { "epoch": 2.381399624315575, "grad_norm": 0.14324088394641876, "learning_rate": 1.677270337403894e-05, "loss": 0.0309, "step": 119170 }, { "epoch": 2.3815994564565766, "grad_norm": 0.13522280752658844, "learning_rate": 1.6764981889724385e-05, "loss": 0.0328, "step": 119180 }, { "epoch": 2.381799288597578, "grad_norm": 0.17806841433048248, "learning_rate": 1.6757261825148746e-05, "loss": 0.028, "step": 119190 }, { "epoch": 2.3819991207385796, "grad_norm": 0.21074986457824707, "learning_rate": 1.6749543180641814e-05, "loss": 0.0266, "step": 119200 }, { "epoch": 2.3821989528795813, "grad_norm": 0.169849693775177, "learning_rate": 1.6741825956533312e-05, "loss": 0.0315, "step": 119210 }, { "epoch": 2.3823987850205826, "grad_norm": 0.11703737825155258, "learning_rate": 1.673411015315291e-05, "loss": 0.0272, "step": 119220 }, { "epoch": 2.3825986171615843, "grad_norm": 0.16443872451782227, "learning_rate": 1.672639577083021e-05, "loss": 0.0235, "step": 119230 }, { "epoch": 2.382798449302586, "grad_norm": 0.15037953853607178, "learning_rate": 1.6718682809894754e-05, "loss": 0.022, "step": 119240 }, { "epoch": 2.3829982814435873, "grad_norm": 0.09277079999446869, "learning_rate": 1.6710971270676033e-05, "loss": 0.0211, "step": 119250 }, { "epoch": 2.383198113584589, "grad_norm": 0.1238044947385788, "learning_rate": 1.6703261153503464e-05, "loss": 0.0239, "step": 119260 }, { "epoch": 2.3833979457255907, "grad_norm": 0.08291785418987274, "learning_rate": 1.6695552458706425e-05, "loss": 0.0181, "step": 119270 }, { "epoch": 2.383597777866592, "grad_norm": 0.12179788202047348, "learning_rate": 1.6687845186614192e-05, "loss": 0.0285, "step": 119280 }, { "epoch": 2.3837976100075937, "grad_norm": 0.11321308463811874, "learning_rate": 1.6680139337555995e-05, "loss": 0.0299, "step": 119290 }, { "epoch": 2.383997442148595, "grad_norm": 0.11095260828733444, "learning_rate": 1.6672434911861067e-05, "loss": 0.0238, "step": 119300 }, { "epoch": 2.3841972742895967, "grad_norm": 0.11855924129486084, "learning_rate": 1.6664731909858483e-05, "loss": 0.0218, "step": 119310 }, { "epoch": 2.3843971064305984, "grad_norm": 0.08155936002731323, "learning_rate": 1.665703033187731e-05, "loss": 0.026, "step": 119320 }, { "epoch": 2.3845969385715997, "grad_norm": 0.1029055044054985, "learning_rate": 1.6649330178246546e-05, "loss": 0.0246, "step": 119330 }, { "epoch": 2.3847967707126014, "grad_norm": 0.10356663912534714, "learning_rate": 1.6641631449295137e-05, "loss": 0.0242, "step": 119340 }, { "epoch": 2.384996602853603, "grad_norm": 0.15234997868537903, "learning_rate": 1.6633934145351948e-05, "loss": 0.0334, "step": 119350 }, { "epoch": 2.3851964349946044, "grad_norm": 0.08683177083730698, "learning_rate": 1.662623826674579e-05, "loss": 0.0184, "step": 119360 }, { "epoch": 2.385396267135606, "grad_norm": 0.2155318409204483, "learning_rate": 1.661854381380545e-05, "loss": 0.0231, "step": 119370 }, { "epoch": 2.385596099276608, "grad_norm": 0.09482292830944061, "learning_rate": 1.6610850786859554e-05, "loss": 0.0248, "step": 119380 }, { "epoch": 2.385795931417609, "grad_norm": 0.11154445260763168, "learning_rate": 1.6603159186236794e-05, "loss": 0.0297, "step": 119390 }, { "epoch": 2.385995763558611, "grad_norm": 0.1346375048160553, "learning_rate": 1.659546901226574e-05, "loss": 0.0243, "step": 119400 }, { "epoch": 2.386195595699612, "grad_norm": 0.12698382139205933, "learning_rate": 1.6587780265274865e-05, "loss": 0.0289, "step": 119410 }, { "epoch": 2.386395427840614, "grad_norm": 0.13835573196411133, "learning_rate": 1.6580092945592622e-05, "loss": 0.0307, "step": 119420 }, { "epoch": 2.3865952599816156, "grad_norm": 0.09322968125343323, "learning_rate": 1.6572407053547444e-05, "loss": 0.0207, "step": 119430 }, { "epoch": 2.386795092122617, "grad_norm": 0.1366145759820938, "learning_rate": 1.6564722589467617e-05, "loss": 0.0299, "step": 119440 }, { "epoch": 2.3869949242636186, "grad_norm": 0.14630630612373352, "learning_rate": 1.6557039553681413e-05, "loss": 0.0236, "step": 119450 }, { "epoch": 2.3871947564046203, "grad_norm": 0.11004027724266052, "learning_rate": 1.654935794651705e-05, "loss": 0.029, "step": 119460 }, { "epoch": 2.3873945885456216, "grad_norm": 0.1139606386423111, "learning_rate": 1.6541677768302656e-05, "loss": 0.0251, "step": 119470 }, { "epoch": 2.3875944206866233, "grad_norm": 0.12765224277973175, "learning_rate": 1.6533999019366326e-05, "loss": 0.0196, "step": 119480 }, { "epoch": 2.387794252827625, "grad_norm": 0.11104617267847061, "learning_rate": 1.652632170003608e-05, "loss": 0.027, "step": 119490 }, { "epoch": 2.3879940849686263, "grad_norm": 0.12959475815296173, "learning_rate": 1.6518645810639893e-05, "loss": 0.0329, "step": 119500 }, { "epoch": 2.388193917109628, "grad_norm": 0.1435183882713318, "learning_rate": 1.6510971351505618e-05, "loss": 0.0186, "step": 119510 }, { "epoch": 2.3883937492506293, "grad_norm": 0.11585313826799393, "learning_rate": 1.6503298322961143e-05, "loss": 0.0279, "step": 119520 }, { "epoch": 2.388593581391631, "grad_norm": 0.07987651228904724, "learning_rate": 1.6495626725334224e-05, "loss": 0.0202, "step": 119530 }, { "epoch": 2.3887934135326327, "grad_norm": 0.11547301709651947, "learning_rate": 1.64879565589526e-05, "loss": 0.0337, "step": 119540 }, { "epoch": 2.388993245673634, "grad_norm": 0.10452867299318314, "learning_rate": 1.6480287824143874e-05, "loss": 0.0329, "step": 119550 }, { "epoch": 2.3891930778146357, "grad_norm": 0.17894987761974335, "learning_rate": 1.6472620521235694e-05, "loss": 0.0325, "step": 119560 }, { "epoch": 2.389392909955637, "grad_norm": 0.10201267153024673, "learning_rate": 1.646495465055559e-05, "loss": 0.0253, "step": 119570 }, { "epoch": 2.3895927420966387, "grad_norm": 0.1738101840019226, "learning_rate": 1.6457290212431e-05, "loss": 0.0254, "step": 119580 }, { "epoch": 2.3897925742376405, "grad_norm": 0.11208481341600418, "learning_rate": 1.6449627207189354e-05, "loss": 0.0261, "step": 119590 }, { "epoch": 2.389992406378642, "grad_norm": 0.11094247549772263, "learning_rate": 1.6441965635157996e-05, "loss": 0.0215, "step": 119600 }, { "epoch": 2.3901922385196435, "grad_norm": 0.11743035167455673, "learning_rate": 1.643430549666422e-05, "loss": 0.025, "step": 119610 }, { "epoch": 2.390392070660645, "grad_norm": 0.12328395992517471, "learning_rate": 1.642664679203525e-05, "loss": 0.031, "step": 119620 }, { "epoch": 2.3905919028016465, "grad_norm": 0.10413390398025513, "learning_rate": 1.6418989521598254e-05, "loss": 0.028, "step": 119630 }, { "epoch": 2.390791734942648, "grad_norm": 0.09933437407016754, "learning_rate": 1.6411333685680337e-05, "loss": 0.0195, "step": 119640 }, { "epoch": 2.39099156708365, "grad_norm": 0.13030067086219788, "learning_rate": 1.6403679284608535e-05, "loss": 0.0336, "step": 119650 }, { "epoch": 2.391191399224651, "grad_norm": 0.16762270033359528, "learning_rate": 1.6396026318709846e-05, "loss": 0.0296, "step": 119660 }, { "epoch": 2.391391231365653, "grad_norm": 0.10040619224309921, "learning_rate": 1.6388374788311188e-05, "loss": 0.0316, "step": 119670 }, { "epoch": 2.391591063506654, "grad_norm": 0.11524117738008499, "learning_rate": 1.6380724693739397e-05, "loss": 0.0218, "step": 119680 }, { "epoch": 2.391790895647656, "grad_norm": 0.1357421725988388, "learning_rate": 1.6373076035321272e-05, "loss": 0.025, "step": 119690 }, { "epoch": 2.3919907277886576, "grad_norm": 0.21433646976947784, "learning_rate": 1.6365428813383592e-05, "loss": 0.0302, "step": 119700 }, { "epoch": 2.392190559929659, "grad_norm": 0.1361864060163498, "learning_rate": 1.6357783028252992e-05, "loss": 0.0267, "step": 119710 }, { "epoch": 2.3923903920706606, "grad_norm": 0.11511655896902084, "learning_rate": 1.63501386802561e-05, "loss": 0.0294, "step": 119720 }, { "epoch": 2.3925902242116623, "grad_norm": 0.14845404028892517, "learning_rate": 1.6342495769719467e-05, "loss": 0.0208, "step": 119730 }, { "epoch": 2.3927900563526636, "grad_norm": 0.11860766261816025, "learning_rate": 1.633485429696958e-05, "loss": 0.0356, "step": 119740 }, { "epoch": 2.3929898884936653, "grad_norm": 0.12076199054718018, "learning_rate": 1.6327214262332873e-05, "loss": 0.0305, "step": 119750 }, { "epoch": 2.393189720634667, "grad_norm": 0.09602417051792145, "learning_rate": 1.6319575666135716e-05, "loss": 0.0223, "step": 119760 }, { "epoch": 2.3933895527756683, "grad_norm": 0.1403690129518509, "learning_rate": 1.6311938508704433e-05, "loss": 0.0279, "step": 119770 }, { "epoch": 2.39358938491667, "grad_norm": 0.11765596270561218, "learning_rate": 1.630430279036521e-05, "loss": 0.0255, "step": 119780 }, { "epoch": 2.3937892170576713, "grad_norm": 0.08863973617553711, "learning_rate": 1.6296668511444296e-05, "loss": 0.0318, "step": 119790 }, { "epoch": 2.393989049198673, "grad_norm": 0.15522272884845734, "learning_rate": 1.62890356722678e-05, "loss": 0.0237, "step": 119800 }, { "epoch": 2.394188881339675, "grad_norm": 0.13166849315166473, "learning_rate": 1.6281404273161753e-05, "loss": 0.026, "step": 119810 }, { "epoch": 2.394388713480676, "grad_norm": 0.10243489593267441, "learning_rate": 1.6273774314452154e-05, "loss": 0.0211, "step": 119820 }, { "epoch": 2.3945885456216778, "grad_norm": 0.18912002444267273, "learning_rate": 1.6266145796465e-05, "loss": 0.019, "step": 119830 }, { "epoch": 2.3947883777626795, "grad_norm": 0.1045553907752037, "learning_rate": 1.62585187195261e-05, "loss": 0.0232, "step": 119840 }, { "epoch": 2.3949882099036808, "grad_norm": 0.122111015021801, "learning_rate": 1.6250893083961305e-05, "loss": 0.0232, "step": 119850 }, { "epoch": 2.3951880420446825, "grad_norm": 0.13647650182247162, "learning_rate": 1.6243268890096353e-05, "loss": 0.0334, "step": 119860 }, { "epoch": 2.395387874185684, "grad_norm": 0.1059805154800415, "learning_rate": 1.6235646138256943e-05, "loss": 0.0287, "step": 119870 }, { "epoch": 2.3955877063266855, "grad_norm": 0.13121984899044037, "learning_rate": 1.62280248287687e-05, "loss": 0.0231, "step": 119880 }, { "epoch": 2.395787538467687, "grad_norm": 0.11164046823978424, "learning_rate": 1.6220404961957197e-05, "loss": 0.0222, "step": 119890 }, { "epoch": 2.3959873706086885, "grad_norm": 0.06699584424495697, "learning_rate": 1.6212786538147956e-05, "loss": 0.0254, "step": 119900 }, { "epoch": 2.39618720274969, "grad_norm": 0.13780459761619568, "learning_rate": 1.620516955766637e-05, "loss": 0.0195, "step": 119910 }, { "epoch": 2.396387034890692, "grad_norm": 0.10280323773622513, "learning_rate": 1.6197554020837874e-05, "loss": 0.0273, "step": 119920 }, { "epoch": 2.396586867031693, "grad_norm": 0.11330006271600723, "learning_rate": 1.6189939927987775e-05, "loss": 0.0218, "step": 119930 }, { "epoch": 2.396786699172695, "grad_norm": 0.11719458550214767, "learning_rate": 1.618232727944135e-05, "loss": 0.0235, "step": 119940 }, { "epoch": 2.3969865313136967, "grad_norm": 0.10413031280040741, "learning_rate": 1.6174716075523755e-05, "loss": 0.0213, "step": 119950 }, { "epoch": 2.397186363454698, "grad_norm": 0.09545854479074478, "learning_rate": 1.616710631656014e-05, "loss": 0.0245, "step": 119960 }, { "epoch": 2.3973861955956997, "grad_norm": 0.13163506984710693, "learning_rate": 1.6159498002875624e-05, "loss": 0.025, "step": 119970 }, { "epoch": 2.3975860277367014, "grad_norm": 0.11318454891443253, "learning_rate": 1.6151891134795165e-05, "loss": 0.0237, "step": 119980 }, { "epoch": 2.3977858598777027, "grad_norm": 0.13453446328639984, "learning_rate": 1.6144285712643736e-05, "loss": 0.0181, "step": 119990 }, { "epoch": 2.3979856920187044, "grad_norm": 0.10821308940649033, "learning_rate": 1.6136681736746227e-05, "loss": 0.0233, "step": 120000 }, { "epoch": 2.3981855241597057, "grad_norm": 0.19784897565841675, "learning_rate": 1.6129079207427466e-05, "loss": 0.0238, "step": 120010 }, { "epoch": 2.3983853563007074, "grad_norm": 0.15725353360176086, "learning_rate": 1.612147812501222e-05, "loss": 0.0323, "step": 120020 }, { "epoch": 2.398585188441709, "grad_norm": 0.12048251926898956, "learning_rate": 1.611387848982519e-05, "loss": 0.0298, "step": 120030 }, { "epoch": 2.3987850205827104, "grad_norm": 0.11195637285709381, "learning_rate": 1.610628030219104e-05, "loss": 0.0258, "step": 120040 }, { "epoch": 2.398984852723712, "grad_norm": 0.12714236974716187, "learning_rate": 1.609868356243429e-05, "loss": 0.0308, "step": 120050 }, { "epoch": 2.399184684864714, "grad_norm": 0.12399640679359436, "learning_rate": 1.6091088270879523e-05, "loss": 0.0321, "step": 120060 }, { "epoch": 2.399384517005715, "grad_norm": 0.10070513188838959, "learning_rate": 1.6083494427851188e-05, "loss": 0.0285, "step": 120070 }, { "epoch": 2.399584349146717, "grad_norm": 0.13452568650245667, "learning_rate": 1.6075902033673644e-05, "loss": 0.021, "step": 120080 }, { "epoch": 2.3997841812877185, "grad_norm": 0.0892396941781044, "learning_rate": 1.6068311088671234e-05, "loss": 0.0263, "step": 120090 }, { "epoch": 2.39998401342872, "grad_norm": 0.11376077681779861, "learning_rate": 1.6060721593168267e-05, "loss": 0.0267, "step": 120100 }, { "epoch": 2.4001838455697215, "grad_norm": 0.1379113495349884, "learning_rate": 1.6053133547488915e-05, "loss": 0.0214, "step": 120110 }, { "epoch": 2.400383677710723, "grad_norm": 0.09619157016277313, "learning_rate": 1.604554695195733e-05, "loss": 0.0195, "step": 120120 }, { "epoch": 2.4005835098517245, "grad_norm": 0.12662802636623383, "learning_rate": 1.6037961806897607e-05, "loss": 0.0348, "step": 120130 }, { "epoch": 2.4007833419927262, "grad_norm": 0.10612378269433975, "learning_rate": 1.603037811263376e-05, "loss": 0.0287, "step": 120140 }, { "epoch": 2.4009831741337275, "grad_norm": 0.10393036156892776, "learning_rate": 1.602279586948976e-05, "loss": 0.0182, "step": 120150 }, { "epoch": 2.4011830062747292, "grad_norm": 0.0879109799861908, "learning_rate": 1.6015215077789496e-05, "loss": 0.0223, "step": 120160 }, { "epoch": 2.401382838415731, "grad_norm": 0.1313178539276123, "learning_rate": 1.6007635737856836e-05, "loss": 0.0322, "step": 120170 }, { "epoch": 2.4015826705567322, "grad_norm": 0.14450249075889587, "learning_rate": 1.6000057850015494e-05, "loss": 0.0237, "step": 120180 }, { "epoch": 2.401782502697734, "grad_norm": 0.15924815833568573, "learning_rate": 1.599248141458924e-05, "loss": 0.021, "step": 120190 }, { "epoch": 2.4019823348387357, "grad_norm": 0.14683333039283752, "learning_rate": 1.598490643190172e-05, "loss": 0.0284, "step": 120200 }, { "epoch": 2.402182166979737, "grad_norm": 0.14008164405822754, "learning_rate": 1.5977332902276498e-05, "loss": 0.0243, "step": 120210 }, { "epoch": 2.4023819991207387, "grad_norm": 0.13686101138591766, "learning_rate": 1.5969760826037106e-05, "loss": 0.0234, "step": 120220 }, { "epoch": 2.40258183126174, "grad_norm": 0.09868216514587402, "learning_rate": 1.596219020350702e-05, "loss": 0.0294, "step": 120230 }, { "epoch": 2.4027816634027417, "grad_norm": 0.1067027598619461, "learning_rate": 1.5954621035009636e-05, "loss": 0.022, "step": 120240 }, { "epoch": 2.4029814955437434, "grad_norm": 0.13820333778858185, "learning_rate": 1.5947053320868305e-05, "loss": 0.0266, "step": 120250 }, { "epoch": 2.4031813276847447, "grad_norm": 0.1218109205365181, "learning_rate": 1.5939487061406294e-05, "loss": 0.0211, "step": 120260 }, { "epoch": 2.4033811598257464, "grad_norm": 0.0875701755285263, "learning_rate": 1.5931922256946824e-05, "loss": 0.031, "step": 120270 }, { "epoch": 2.4035809919667477, "grad_norm": 0.1037156879901886, "learning_rate": 1.5924358907813048e-05, "loss": 0.0373, "step": 120280 }, { "epoch": 2.4037808241077494, "grad_norm": 0.1307826191186905, "learning_rate": 1.5916797014328065e-05, "loss": 0.0201, "step": 120290 }, { "epoch": 2.403980656248751, "grad_norm": 0.1362922489643097, "learning_rate": 1.5909236576814917e-05, "loss": 0.0262, "step": 120300 }, { "epoch": 2.404180488389753, "grad_norm": 0.08662594854831696, "learning_rate": 1.5901677595596537e-05, "loss": 0.0239, "step": 120310 }, { "epoch": 2.404380320530754, "grad_norm": 0.1064356192946434, "learning_rate": 1.5894120070995834e-05, "loss": 0.0228, "step": 120320 }, { "epoch": 2.404580152671756, "grad_norm": 0.11231851577758789, "learning_rate": 1.588656400333568e-05, "loss": 0.0248, "step": 120330 }, { "epoch": 2.404779984812757, "grad_norm": 0.12470430880784988, "learning_rate": 1.5879009392938865e-05, "loss": 0.0182, "step": 120340 }, { "epoch": 2.404979816953759, "grad_norm": 0.12426722049713135, "learning_rate": 1.587145624012807e-05, "loss": 0.0294, "step": 120350 }, { "epoch": 2.4051796490947606, "grad_norm": 0.11926772445440292, "learning_rate": 1.586390454522595e-05, "loss": 0.0317, "step": 120360 }, { "epoch": 2.405379481235762, "grad_norm": 0.11261766403913498, "learning_rate": 1.585635430855515e-05, "loss": 0.0297, "step": 120370 }, { "epoch": 2.4055793133767636, "grad_norm": 0.13472066819667816, "learning_rate": 1.5848805530438154e-05, "loss": 0.0334, "step": 120380 }, { "epoch": 2.405779145517765, "grad_norm": 0.13246752321720123, "learning_rate": 1.5841258211197445e-05, "loss": 0.02, "step": 120390 }, { "epoch": 2.4059789776587666, "grad_norm": 0.1273224651813507, "learning_rate": 1.583371235115544e-05, "loss": 0.0237, "step": 120400 }, { "epoch": 2.4061788097997683, "grad_norm": 0.15643751621246338, "learning_rate": 1.5826167950634473e-05, "loss": 0.0327, "step": 120410 }, { "epoch": 2.40637864194077, "grad_norm": 0.180513396859169, "learning_rate": 1.5818625009956835e-05, "loss": 0.021, "step": 120420 }, { "epoch": 2.4065784740817713, "grad_norm": 0.10855641961097717, "learning_rate": 1.5811083529444742e-05, "loss": 0.0287, "step": 120430 }, { "epoch": 2.406778306222773, "grad_norm": 0.12069126218557358, "learning_rate": 1.580354350942037e-05, "loss": 0.0263, "step": 120440 }, { "epoch": 2.4069781383637743, "grad_norm": 0.11361788958311081, "learning_rate": 1.579600495020577e-05, "loss": 0.0176, "step": 120450 }, { "epoch": 2.407177970504776, "grad_norm": 0.1845761388540268, "learning_rate": 1.578846785212301e-05, "loss": 0.0245, "step": 120460 }, { "epoch": 2.4073778026457777, "grad_norm": 0.11529770493507385, "learning_rate": 1.5780932215494076e-05, "loss": 0.025, "step": 120470 }, { "epoch": 2.407577634786779, "grad_norm": 0.1351000964641571, "learning_rate": 1.5773398040640842e-05, "loss": 0.0282, "step": 120480 }, { "epoch": 2.4077774669277807, "grad_norm": 0.13025768101215363, "learning_rate": 1.5765865327885165e-05, "loss": 0.0324, "step": 120490 }, { "epoch": 2.407977299068782, "grad_norm": 0.1829046756029129, "learning_rate": 1.5758334077548832e-05, "loss": 0.0357, "step": 120500 }, { "epoch": 2.4081771312097837, "grad_norm": 0.12592247128486633, "learning_rate": 1.5750804289953563e-05, "loss": 0.0272, "step": 120510 }, { "epoch": 2.4083769633507854, "grad_norm": 0.1190117672085762, "learning_rate": 1.5743275965421018e-05, "loss": 0.0272, "step": 120520 }, { "epoch": 2.4085767954917867, "grad_norm": 0.11177245527505875, "learning_rate": 1.573574910427279e-05, "loss": 0.0228, "step": 120530 }, { "epoch": 2.4087766276327884, "grad_norm": 0.15441961586475372, "learning_rate": 1.5728223706830413e-05, "loss": 0.0317, "step": 120540 }, { "epoch": 2.40897645977379, "grad_norm": 0.14266906678676605, "learning_rate": 1.5720699773415366e-05, "loss": 0.0241, "step": 120550 }, { "epoch": 2.4091762919147914, "grad_norm": 0.13151448965072632, "learning_rate": 1.5713177304349046e-05, "loss": 0.0264, "step": 120560 }, { "epoch": 2.409376124055793, "grad_norm": 0.1514711230993271, "learning_rate": 1.5705656299952826e-05, "loss": 0.0262, "step": 120570 }, { "epoch": 2.409575956196795, "grad_norm": 0.11570199579000473, "learning_rate": 1.569813676054795e-05, "loss": 0.0205, "step": 120580 }, { "epoch": 2.409775788337796, "grad_norm": 0.105253204703331, "learning_rate": 1.5690618686455637e-05, "loss": 0.0265, "step": 120590 }, { "epoch": 2.409975620478798, "grad_norm": 0.14061839878559113, "learning_rate": 1.5683102077997102e-05, "loss": 0.0337, "step": 120600 }, { "epoch": 2.410175452619799, "grad_norm": 0.1430250108242035, "learning_rate": 1.5675586935493387e-05, "loss": 0.0261, "step": 120610 }, { "epoch": 2.410375284760801, "grad_norm": 0.12032145261764526, "learning_rate": 1.5668073259265536e-05, "loss": 0.0266, "step": 120620 }, { "epoch": 2.4105751169018026, "grad_norm": 0.12529611587524414, "learning_rate": 1.5660561049634535e-05, "loss": 0.0196, "step": 120630 }, { "epoch": 2.410774949042804, "grad_norm": 0.10545129328966141, "learning_rate": 1.565305030692128e-05, "loss": 0.0191, "step": 120640 }, { "epoch": 2.4109747811838056, "grad_norm": 0.12183381617069244, "learning_rate": 1.5645541031446613e-05, "loss": 0.0189, "step": 120650 }, { "epoch": 2.4111746133248073, "grad_norm": 0.09009584039449692, "learning_rate": 1.563803322353133e-05, "loss": 0.0257, "step": 120660 }, { "epoch": 2.4113744454658086, "grad_norm": 0.12555542588233948, "learning_rate": 1.5630526883496154e-05, "loss": 0.0218, "step": 120670 }, { "epoch": 2.4115742776068103, "grad_norm": 0.1293715238571167, "learning_rate": 1.56230220116617e-05, "loss": 0.0279, "step": 120680 }, { "epoch": 2.411774109747812, "grad_norm": 0.1708258092403412, "learning_rate": 1.5615518608348606e-05, "loss": 0.0307, "step": 120690 }, { "epoch": 2.4119739418888133, "grad_norm": 0.14011472463607788, "learning_rate": 1.560801667387741e-05, "loss": 0.0221, "step": 120700 }, { "epoch": 2.412173774029815, "grad_norm": 0.1287684142589569, "learning_rate": 1.5600516208568546e-05, "loss": 0.0266, "step": 120710 }, { "epoch": 2.4123736061708163, "grad_norm": 0.17501132190227509, "learning_rate": 1.5593017212742416e-05, "loss": 0.0208, "step": 120720 }, { "epoch": 2.412573438311818, "grad_norm": 0.07770247757434845, "learning_rate": 1.5585519686719403e-05, "loss": 0.0211, "step": 120730 }, { "epoch": 2.4127732704528198, "grad_norm": 0.1125229001045227, "learning_rate": 1.5578023630819783e-05, "loss": 0.0191, "step": 120740 }, { "epoch": 2.412973102593821, "grad_norm": 0.09451109170913696, "learning_rate": 1.557052904536374e-05, "loss": 0.0196, "step": 120750 }, { "epoch": 2.4131729347348227, "grad_norm": 0.09779219329357147, "learning_rate": 1.5563035930671432e-05, "loss": 0.0193, "step": 120760 }, { "epoch": 2.4133727668758245, "grad_norm": 0.11469442397356033, "learning_rate": 1.5555544287063e-05, "loss": 0.0266, "step": 120770 }, { "epoch": 2.4135725990168257, "grad_norm": 0.16734591126441956, "learning_rate": 1.5548054114858425e-05, "loss": 0.0302, "step": 120780 }, { "epoch": 2.4137724311578275, "grad_norm": 0.09895502775907516, "learning_rate": 1.5540565414377682e-05, "loss": 0.0257, "step": 120790 }, { "epoch": 2.413972263298829, "grad_norm": 0.15761765837669373, "learning_rate": 1.5533078185940685e-05, "loss": 0.0207, "step": 120800 }, { "epoch": 2.4141720954398305, "grad_norm": 0.11328139156103134, "learning_rate": 1.5525592429867277e-05, "loss": 0.0196, "step": 120810 }, { "epoch": 2.414371927580832, "grad_norm": 0.12054566293954849, "learning_rate": 1.5518108146477216e-05, "loss": 0.0266, "step": 120820 }, { "epoch": 2.4145717597218335, "grad_norm": 0.13271121680736542, "learning_rate": 1.5510625336090234e-05, "loss": 0.0276, "step": 120830 }, { "epoch": 2.414771591862835, "grad_norm": 0.14026641845703125, "learning_rate": 1.5503143999025998e-05, "loss": 0.0289, "step": 120840 }, { "epoch": 2.414971424003837, "grad_norm": 0.1870831996202469, "learning_rate": 1.5495664135604042e-05, "loss": 0.0203, "step": 120850 }, { "epoch": 2.415171256144838, "grad_norm": 0.10977309942245483, "learning_rate": 1.5488185746143945e-05, "loss": 0.025, "step": 120860 }, { "epoch": 2.41537108828584, "grad_norm": 0.13521459698677063, "learning_rate": 1.5480708830965167e-05, "loss": 0.0252, "step": 120870 }, { "epoch": 2.4155709204268416, "grad_norm": 0.1616000384092331, "learning_rate": 1.547323339038708e-05, "loss": 0.0198, "step": 120880 }, { "epoch": 2.415770752567843, "grad_norm": 0.10643645375967026, "learning_rate": 1.5465759424729037e-05, "loss": 0.0156, "step": 120890 }, { "epoch": 2.4159705847088446, "grad_norm": 0.09220104664564133, "learning_rate": 1.5458286934310306e-05, "loss": 0.0252, "step": 120900 }, { "epoch": 2.4161704168498463, "grad_norm": 0.10736657679080963, "learning_rate": 1.5450815919450102e-05, "loss": 0.0224, "step": 120910 }, { "epoch": 2.4163702489908476, "grad_norm": 0.11362139880657196, "learning_rate": 1.5443346380467577e-05, "loss": 0.035, "step": 120920 }, { "epoch": 2.4165700811318493, "grad_norm": 0.09550857543945312, "learning_rate": 1.5435878317681806e-05, "loss": 0.0268, "step": 120930 }, { "epoch": 2.4167699132728506, "grad_norm": 0.1059604361653328, "learning_rate": 1.5428411731411825e-05, "loss": 0.0287, "step": 120940 }, { "epoch": 2.4169697454138523, "grad_norm": 0.11236449331045151, "learning_rate": 1.5420946621976577e-05, "loss": 0.0305, "step": 120950 }, { "epoch": 2.417169577554854, "grad_norm": 0.16243772208690643, "learning_rate": 1.5413482989694968e-05, "loss": 0.0231, "step": 120960 }, { "epoch": 2.4173694096958553, "grad_norm": 0.1262081265449524, "learning_rate": 1.5406020834885844e-05, "loss": 0.0254, "step": 120970 }, { "epoch": 2.417569241836857, "grad_norm": 0.13103514909744263, "learning_rate": 1.5398560157867943e-05, "loss": 0.0316, "step": 120980 }, { "epoch": 2.417769073977859, "grad_norm": 0.11077424138784409, "learning_rate": 1.5391100958959975e-05, "loss": 0.0295, "step": 120990 }, { "epoch": 2.41796890611886, "grad_norm": 0.14366024732589722, "learning_rate": 1.5383643238480622e-05, "loss": 0.0205, "step": 121000 }, { "epoch": 2.418168738259862, "grad_norm": 0.10825921595096588, "learning_rate": 1.5376186996748425e-05, "loss": 0.0198, "step": 121010 }, { "epoch": 2.4183685704008635, "grad_norm": 0.11684183776378632, "learning_rate": 1.536873223408191e-05, "loss": 0.0296, "step": 121020 }, { "epoch": 2.418568402541865, "grad_norm": 0.09776389598846436, "learning_rate": 1.536127895079954e-05, "loss": 0.0236, "step": 121030 }, { "epoch": 2.4187682346828665, "grad_norm": 0.1331098973751068, "learning_rate": 1.5353827147219696e-05, "loss": 0.0245, "step": 121040 }, { "epoch": 2.418968066823868, "grad_norm": 0.09750299155712128, "learning_rate": 1.5346376823660706e-05, "loss": 0.0215, "step": 121050 }, { "epoch": 2.4191678989648695, "grad_norm": 0.0792793557047844, "learning_rate": 1.5338927980440843e-05, "loss": 0.0306, "step": 121060 }, { "epoch": 2.419367731105871, "grad_norm": 0.11329212039709091, "learning_rate": 1.5331480617878313e-05, "loss": 0.0333, "step": 121070 }, { "epoch": 2.4195675632468725, "grad_norm": 0.1076827421784401, "learning_rate": 1.5324034736291217e-05, "loss": 0.0226, "step": 121080 }, { "epoch": 2.419767395387874, "grad_norm": 0.10929916799068451, "learning_rate": 1.5316590335997666e-05, "loss": 0.0228, "step": 121090 }, { "epoch": 2.4199672275288755, "grad_norm": 0.10140521079301834, "learning_rate": 1.5309147417315668e-05, "loss": 0.0272, "step": 121100 }, { "epoch": 2.420167059669877, "grad_norm": 0.1525951325893402, "learning_rate": 1.5301705980563168e-05, "loss": 0.022, "step": 121110 }, { "epoch": 2.420366891810879, "grad_norm": 0.14218805730342865, "learning_rate": 1.5294266026058024e-05, "loss": 0.0205, "step": 121120 }, { "epoch": 2.4205667239518807, "grad_norm": 0.1274704784154892, "learning_rate": 1.528682755411809e-05, "loss": 0.0254, "step": 121130 }, { "epoch": 2.420766556092882, "grad_norm": 0.15009331703186035, "learning_rate": 1.527939056506113e-05, "loss": 0.0278, "step": 121140 }, { "epoch": 2.4209663882338837, "grad_norm": 0.08611800521612167, "learning_rate": 1.52719550592048e-05, "loss": 0.0196, "step": 121150 }, { "epoch": 2.421166220374885, "grad_norm": 0.0780661404132843, "learning_rate": 1.5264521036866757e-05, "loss": 0.0217, "step": 121160 }, { "epoch": 2.4213660525158867, "grad_norm": 0.12781089544296265, "learning_rate": 1.525708849836457e-05, "loss": 0.0259, "step": 121170 }, { "epoch": 2.4215658846568884, "grad_norm": 0.1023193821310997, "learning_rate": 1.524965744401573e-05, "loss": 0.0357, "step": 121180 }, { "epoch": 2.4217657167978897, "grad_norm": 0.11853814870119095, "learning_rate": 1.5242227874137688e-05, "loss": 0.0225, "step": 121190 }, { "epoch": 2.4219655489388914, "grad_norm": 0.14811719954013824, "learning_rate": 1.5234799789047815e-05, "loss": 0.0235, "step": 121200 }, { "epoch": 2.4221653810798927, "grad_norm": 0.13634532690048218, "learning_rate": 1.5227373189063438e-05, "loss": 0.0239, "step": 121210 }, { "epoch": 2.4223652132208944, "grad_norm": 0.08854053169488907, "learning_rate": 1.5219948074501789e-05, "loss": 0.0214, "step": 121220 }, { "epoch": 2.422565045361896, "grad_norm": 0.1591663807630539, "learning_rate": 1.5212524445680071e-05, "loss": 0.0308, "step": 121230 }, { "epoch": 2.4227648775028974, "grad_norm": 0.11989231407642365, "learning_rate": 1.520510230291542e-05, "loss": 0.0219, "step": 121240 }, { "epoch": 2.422964709643899, "grad_norm": 0.132919043302536, "learning_rate": 1.5197681646524858e-05, "loss": 0.0234, "step": 121250 }, { "epoch": 2.423164541784901, "grad_norm": 0.12506286799907684, "learning_rate": 1.5190262476825384e-05, "loss": 0.0231, "step": 121260 }, { "epoch": 2.423364373925902, "grad_norm": 0.10272027552127838, "learning_rate": 1.5182844794133988e-05, "loss": 0.0254, "step": 121270 }, { "epoch": 2.423564206066904, "grad_norm": 0.09758538752794266, "learning_rate": 1.5175428598767482e-05, "loss": 0.0229, "step": 121280 }, { "epoch": 2.4237640382079055, "grad_norm": 0.1204763799905777, "learning_rate": 1.5168013891042692e-05, "loss": 0.0337, "step": 121290 }, { "epoch": 2.423963870348907, "grad_norm": 0.1074548065662384, "learning_rate": 1.5160600671276359e-05, "loss": 0.0187, "step": 121300 }, { "epoch": 2.4241637024899085, "grad_norm": 0.11246145516633987, "learning_rate": 1.5153188939785167e-05, "loss": 0.0298, "step": 121310 }, { "epoch": 2.42436353463091, "grad_norm": 0.11101752519607544, "learning_rate": 1.514577869688572e-05, "loss": 0.0293, "step": 121320 }, { "epoch": 2.4245633667719115, "grad_norm": 0.10120446234941483, "learning_rate": 1.5138369942894587e-05, "loss": 0.0286, "step": 121330 }, { "epoch": 2.4247631989129133, "grad_norm": 0.12478524446487427, "learning_rate": 1.5130962678128258e-05, "loss": 0.0196, "step": 121340 }, { "epoch": 2.4249630310539145, "grad_norm": 0.10845265537500381, "learning_rate": 1.5123556902903118e-05, "loss": 0.0203, "step": 121350 }, { "epoch": 2.4251628631949163, "grad_norm": 0.08687345683574677, "learning_rate": 1.5116152617535567e-05, "loss": 0.0279, "step": 121360 }, { "epoch": 2.425362695335918, "grad_norm": 0.16041575372219086, "learning_rate": 1.5108749822341917e-05, "loss": 0.0342, "step": 121370 }, { "epoch": 2.4255625274769193, "grad_norm": 0.08297865837812424, "learning_rate": 1.5101348517638353e-05, "loss": 0.0213, "step": 121380 }, { "epoch": 2.425762359617921, "grad_norm": 0.2078120857477188, "learning_rate": 1.5093948703741057e-05, "loss": 0.025, "step": 121390 }, { "epoch": 2.4259621917589227, "grad_norm": 0.10470586270093918, "learning_rate": 1.508655038096618e-05, "loss": 0.0203, "step": 121400 }, { "epoch": 2.426162023899924, "grad_norm": 0.10466762632131577, "learning_rate": 1.5079153549629721e-05, "loss": 0.0259, "step": 121410 }, { "epoch": 2.4263618560409257, "grad_norm": 0.12143702059984207, "learning_rate": 1.5071758210047664e-05, "loss": 0.025, "step": 121420 }, { "epoch": 2.426561688181927, "grad_norm": 0.1486445516347885, "learning_rate": 1.5064364362535932e-05, "loss": 0.0291, "step": 121430 }, { "epoch": 2.4267615203229287, "grad_norm": 0.1315433233976364, "learning_rate": 1.505697200741038e-05, "loss": 0.0233, "step": 121440 }, { "epoch": 2.4269613524639304, "grad_norm": 0.10552793741226196, "learning_rate": 1.5049581144986791e-05, "loss": 0.0205, "step": 121450 }, { "epoch": 2.4271611846049317, "grad_norm": 0.11002405732870102, "learning_rate": 1.5042191775580883e-05, "loss": 0.0243, "step": 121460 }, { "epoch": 2.4273610167459334, "grad_norm": 0.1043272614479065, "learning_rate": 1.5034803899508343e-05, "loss": 0.0232, "step": 121470 }, { "epoch": 2.427560848886935, "grad_norm": 0.13273079693317413, "learning_rate": 1.5027417517084718e-05, "loss": 0.0288, "step": 121480 }, { "epoch": 2.4277606810279364, "grad_norm": 0.14013968408107758, "learning_rate": 1.5020032628625579e-05, "loss": 0.0203, "step": 121490 }, { "epoch": 2.427960513168938, "grad_norm": 0.11143344640731812, "learning_rate": 1.5012649234446397e-05, "loss": 0.0222, "step": 121500 }, { "epoch": 2.42816034530994, "grad_norm": 0.12721633911132812, "learning_rate": 1.5005267334862571e-05, "loss": 0.0209, "step": 121510 }, { "epoch": 2.428360177450941, "grad_norm": 0.11879955977201462, "learning_rate": 1.4997886930189425e-05, "loss": 0.0245, "step": 121520 }, { "epoch": 2.428560009591943, "grad_norm": 0.12815351784229279, "learning_rate": 1.4990508020742233e-05, "loss": 0.0242, "step": 121530 }, { "epoch": 2.428759841732944, "grad_norm": 0.1295326054096222, "learning_rate": 1.4983130606836254e-05, "loss": 0.0284, "step": 121540 }, { "epoch": 2.428959673873946, "grad_norm": 0.11197911947965622, "learning_rate": 1.4975754688786597e-05, "loss": 0.0296, "step": 121550 }, { "epoch": 2.4291595060149476, "grad_norm": 0.11580479145050049, "learning_rate": 1.496838026690835e-05, "loss": 0.0184, "step": 121560 }, { "epoch": 2.429359338155949, "grad_norm": 0.14251808822155, "learning_rate": 1.4961007341516548e-05, "loss": 0.0267, "step": 121570 }, { "epoch": 2.4295591702969506, "grad_norm": 0.13229766488075256, "learning_rate": 1.4953635912926145e-05, "loss": 0.0259, "step": 121580 }, { "epoch": 2.4297590024379523, "grad_norm": 0.1364705115556717, "learning_rate": 1.4946265981452028e-05, "loss": 0.0285, "step": 121590 }, { "epoch": 2.4299588345789536, "grad_norm": 0.09508887678384781, "learning_rate": 1.4938897547409037e-05, "loss": 0.0346, "step": 121600 }, { "epoch": 2.4301586667199553, "grad_norm": 0.15422672033309937, "learning_rate": 1.4931530611111948e-05, "loss": 0.0274, "step": 121610 }, { "epoch": 2.430358498860957, "grad_norm": 0.12191715836524963, "learning_rate": 1.4924165172875416e-05, "loss": 0.0214, "step": 121620 }, { "epoch": 2.4305583310019583, "grad_norm": 0.11633589118719101, "learning_rate": 1.4916801233014127e-05, "loss": 0.0281, "step": 121630 }, { "epoch": 2.43075816314296, "grad_norm": 0.10152653604745865, "learning_rate": 1.4909438791842656e-05, "loss": 0.023, "step": 121640 }, { "epoch": 2.4309579952839613, "grad_norm": 0.14111125469207764, "learning_rate": 1.4902077849675483e-05, "loss": 0.0242, "step": 121650 }, { "epoch": 2.431157827424963, "grad_norm": 0.1250624805688858, "learning_rate": 1.4894718406827045e-05, "loss": 0.024, "step": 121660 }, { "epoch": 2.4313576595659647, "grad_norm": 0.15821970999240875, "learning_rate": 1.4887360463611788e-05, "loss": 0.0295, "step": 121670 }, { "epoch": 2.431557491706966, "grad_norm": 0.12355727702379227, "learning_rate": 1.4880004020343964e-05, "loss": 0.0279, "step": 121680 }, { "epoch": 2.4317573238479677, "grad_norm": 0.0844331681728363, "learning_rate": 1.4872649077337853e-05, "loss": 0.0228, "step": 121690 }, { "epoch": 2.4319571559889694, "grad_norm": 0.12391337752342224, "learning_rate": 1.4865295634907645e-05, "loss": 0.0188, "step": 121700 }, { "epoch": 2.4321569881299707, "grad_norm": 0.12265428155660629, "learning_rate": 1.485794369336746e-05, "loss": 0.0245, "step": 121710 }, { "epoch": 2.4323568202709724, "grad_norm": 0.1110088899731636, "learning_rate": 1.4850593253031358e-05, "loss": 0.0221, "step": 121720 }, { "epoch": 2.432556652411974, "grad_norm": 0.1351398378610611, "learning_rate": 1.484324431421335e-05, "loss": 0.0246, "step": 121730 }, { "epoch": 2.4327564845529754, "grad_norm": 0.13585346937179565, "learning_rate": 1.4835896877227368e-05, "loss": 0.0197, "step": 121740 }, { "epoch": 2.432956316693977, "grad_norm": 0.10251668840646744, "learning_rate": 1.4828550942387248e-05, "loss": 0.018, "step": 121750 }, { "epoch": 2.4331561488349784, "grad_norm": 0.11173032224178314, "learning_rate": 1.482120651000683e-05, "loss": 0.0253, "step": 121760 }, { "epoch": 2.43335598097598, "grad_norm": 0.1387442648410797, "learning_rate": 1.4813863580399862e-05, "loss": 0.0211, "step": 121770 }, { "epoch": 2.433555813116982, "grad_norm": 0.11994656920433044, "learning_rate": 1.4806522153879987e-05, "loss": 0.0216, "step": 121780 }, { "epoch": 2.433755645257983, "grad_norm": 0.18480587005615234, "learning_rate": 1.4799182230760833e-05, "loss": 0.0323, "step": 121790 }, { "epoch": 2.433955477398985, "grad_norm": 0.11940151453018188, "learning_rate": 1.4791843811355949e-05, "loss": 0.0205, "step": 121800 }, { "epoch": 2.4341553095399866, "grad_norm": 0.09932483732700348, "learning_rate": 1.4784506895978817e-05, "loss": 0.0242, "step": 121810 }, { "epoch": 2.434355141680988, "grad_norm": 0.13512374460697174, "learning_rate": 1.477717148494286e-05, "loss": 0.0287, "step": 121820 }, { "epoch": 2.4345549738219896, "grad_norm": 0.21834725141525269, "learning_rate": 1.476983757856143e-05, "loss": 0.026, "step": 121830 }, { "epoch": 2.4347548059629913, "grad_norm": 0.11913025379180908, "learning_rate": 1.4762505177147812e-05, "loss": 0.0227, "step": 121840 }, { "epoch": 2.4349546381039926, "grad_norm": 0.12778577208518982, "learning_rate": 1.475517428101525e-05, "loss": 0.0206, "step": 121850 }, { "epoch": 2.4351544702449943, "grad_norm": 0.10210607945919037, "learning_rate": 1.4747844890476886e-05, "loss": 0.022, "step": 121860 }, { "epoch": 2.4353543023859956, "grad_norm": 0.15034231543540955, "learning_rate": 1.4740517005845849e-05, "loss": 0.0233, "step": 121870 }, { "epoch": 2.4355541345269973, "grad_norm": 0.10526018589735031, "learning_rate": 1.473319062743514e-05, "loss": 0.0213, "step": 121880 }, { "epoch": 2.435753966667999, "grad_norm": 0.10738233476877213, "learning_rate": 1.4725865755557716e-05, "loss": 0.0255, "step": 121890 }, { "epoch": 2.4359537988090003, "grad_norm": 0.14733102917671204, "learning_rate": 1.4718542390526529e-05, "loss": 0.0201, "step": 121900 }, { "epoch": 2.436153630950002, "grad_norm": 0.14348363876342773, "learning_rate": 1.4711220532654408e-05, "loss": 0.0289, "step": 121910 }, { "epoch": 2.4363534630910033, "grad_norm": 0.1322447955608368, "learning_rate": 1.4703900182254104e-05, "loss": 0.0214, "step": 121920 }, { "epoch": 2.436553295232005, "grad_norm": 0.11999547481536865, "learning_rate": 1.4696581339638327e-05, "loss": 0.0235, "step": 121930 }, { "epoch": 2.4367531273730068, "grad_norm": 0.1414339244365692, "learning_rate": 1.4689264005119774e-05, "loss": 0.025, "step": 121940 }, { "epoch": 2.4369529595140085, "grad_norm": 0.08641677349805832, "learning_rate": 1.468194817901098e-05, "loss": 0.0207, "step": 121950 }, { "epoch": 2.4371527916550098, "grad_norm": 0.12346449494361877, "learning_rate": 1.4674633861624476e-05, "loss": 0.0205, "step": 121960 }, { "epoch": 2.4373526237960115, "grad_norm": 0.1032990887761116, "learning_rate": 1.466732105327272e-05, "loss": 0.0273, "step": 121970 }, { "epoch": 2.4375524559370128, "grad_norm": 0.12376835197210312, "learning_rate": 1.4660009754268094e-05, "loss": 0.0175, "step": 121980 }, { "epoch": 2.4377522880780145, "grad_norm": 0.16881993412971497, "learning_rate": 1.4652699964922939e-05, "loss": 0.0205, "step": 121990 }, { "epoch": 2.437952120219016, "grad_norm": 0.09182214736938477, "learning_rate": 1.4645391685549503e-05, "loss": 0.0153, "step": 122000 }, { "epoch": 2.4381519523600175, "grad_norm": 0.159847229719162, "learning_rate": 1.4638084916459999e-05, "loss": 0.0223, "step": 122010 }, { "epoch": 2.438351784501019, "grad_norm": 0.10074866563081741, "learning_rate": 1.4630779657966515e-05, "loss": 0.0177, "step": 122020 }, { "epoch": 2.4385516166420205, "grad_norm": 0.1360667496919632, "learning_rate": 1.4623475910381167e-05, "loss": 0.0205, "step": 122030 }, { "epoch": 2.438751448783022, "grad_norm": 0.10171405971050262, "learning_rate": 1.4616173674015955e-05, "loss": 0.026, "step": 122040 }, { "epoch": 2.438951280924024, "grad_norm": 0.27090340852737427, "learning_rate": 1.4608872949182783e-05, "loss": 0.0314, "step": 122050 }, { "epoch": 2.439151113065025, "grad_norm": 0.13135525584220886, "learning_rate": 1.460157373619353e-05, "loss": 0.0268, "step": 122060 }, { "epoch": 2.439350945206027, "grad_norm": 0.11064846813678741, "learning_rate": 1.4594276035360049e-05, "loss": 0.0259, "step": 122070 }, { "epoch": 2.4395507773470286, "grad_norm": 0.09204462170600891, "learning_rate": 1.4586979846994037e-05, "loss": 0.022, "step": 122080 }, { "epoch": 2.43975060948803, "grad_norm": 0.12991252541542053, "learning_rate": 1.4579685171407187e-05, "loss": 0.0235, "step": 122090 }, { "epoch": 2.4399504416290316, "grad_norm": 0.09031952172517776, "learning_rate": 1.4572392008911123e-05, "loss": 0.0206, "step": 122100 }, { "epoch": 2.4401502737700334, "grad_norm": 0.16202490031719208, "learning_rate": 1.4565100359817386e-05, "loss": 0.0277, "step": 122110 }, { "epoch": 2.4403501059110346, "grad_norm": 0.08858639746904373, "learning_rate": 1.4557810224437463e-05, "loss": 0.0245, "step": 122120 }, { "epoch": 2.4405499380520363, "grad_norm": 0.15307402610778809, "learning_rate": 1.4550521603082784e-05, "loss": 0.0282, "step": 122130 }, { "epoch": 2.4407497701930376, "grad_norm": 0.10904953628778458, "learning_rate": 1.4543234496064711e-05, "loss": 0.0215, "step": 122140 }, { "epoch": 2.4409496023340393, "grad_norm": 0.12769192457199097, "learning_rate": 1.4535948903694497e-05, "loss": 0.0289, "step": 122150 }, { "epoch": 2.441149434475041, "grad_norm": 0.15467673540115356, "learning_rate": 1.4528664826283405e-05, "loss": 0.0301, "step": 122160 }, { "epoch": 2.4413492666160423, "grad_norm": 0.10502341389656067, "learning_rate": 1.4521382264142614e-05, "loss": 0.019, "step": 122170 }, { "epoch": 2.441549098757044, "grad_norm": 0.15146493911743164, "learning_rate": 1.451410121758317e-05, "loss": 0.0229, "step": 122180 }, { "epoch": 2.441748930898046, "grad_norm": 0.09428104013204575, "learning_rate": 1.450682168691614e-05, "loss": 0.0174, "step": 122190 }, { "epoch": 2.441948763039047, "grad_norm": 0.12147969007492065, "learning_rate": 1.449954367245248e-05, "loss": 0.0335, "step": 122200 }, { "epoch": 2.442148595180049, "grad_norm": 0.11891502141952515, "learning_rate": 1.4492267174503098e-05, "loss": 0.0263, "step": 122210 }, { "epoch": 2.4423484273210505, "grad_norm": 0.12494532018899918, "learning_rate": 1.4484992193378837e-05, "loss": 0.0267, "step": 122220 }, { "epoch": 2.442548259462052, "grad_norm": 0.1191871166229248, "learning_rate": 1.4477718729390465e-05, "loss": 0.0275, "step": 122230 }, { "epoch": 2.4427480916030535, "grad_norm": 0.08435172587633133, "learning_rate": 1.447044678284869e-05, "loss": 0.0172, "step": 122240 }, { "epoch": 2.442947923744055, "grad_norm": 0.16371065378189087, "learning_rate": 1.4463176354064161e-05, "loss": 0.0262, "step": 122250 }, { "epoch": 2.4431477558850565, "grad_norm": 0.07279609143733978, "learning_rate": 1.4455907443347455e-05, "loss": 0.0281, "step": 122260 }, { "epoch": 2.4433475880260582, "grad_norm": 0.1072034239768982, "learning_rate": 1.4448640051009099e-05, "loss": 0.0263, "step": 122270 }, { "epoch": 2.4435474201670595, "grad_norm": 0.1013498529791832, "learning_rate": 1.4441374177359518e-05, "loss": 0.0284, "step": 122280 }, { "epoch": 2.4437472523080612, "grad_norm": 0.14719097316265106, "learning_rate": 1.4434109822709091e-05, "loss": 0.0191, "step": 122290 }, { "epoch": 2.443947084449063, "grad_norm": 0.10654503852128983, "learning_rate": 1.442684698736817e-05, "loss": 0.0164, "step": 122300 }, { "epoch": 2.4441469165900642, "grad_norm": 0.11266818642616272, "learning_rate": 1.4419585671647012e-05, "loss": 0.0156, "step": 122310 }, { "epoch": 2.444346748731066, "grad_norm": 0.1518220454454422, "learning_rate": 1.4412325875855775e-05, "loss": 0.0287, "step": 122320 }, { "epoch": 2.4445465808720677, "grad_norm": 0.09206660836935043, "learning_rate": 1.440506760030459e-05, "loss": 0.026, "step": 122330 }, { "epoch": 2.444746413013069, "grad_norm": 0.12245627492666245, "learning_rate": 1.439781084530355e-05, "loss": 0.022, "step": 122340 }, { "epoch": 2.4449462451540707, "grad_norm": 0.12298178672790527, "learning_rate": 1.4390555611162621e-05, "loss": 0.0337, "step": 122350 }, { "epoch": 2.445146077295072, "grad_norm": 0.12431380897760391, "learning_rate": 1.4383301898191731e-05, "loss": 0.0222, "step": 122360 }, { "epoch": 2.4453459094360737, "grad_norm": 0.1354093849658966, "learning_rate": 1.4376049706700755e-05, "loss": 0.0245, "step": 122370 }, { "epoch": 2.4455457415770754, "grad_norm": 0.10553883761167526, "learning_rate": 1.4368799036999498e-05, "loss": 0.0236, "step": 122380 }, { "epoch": 2.4457455737180767, "grad_norm": 0.15396493673324585, "learning_rate": 1.4361549889397685e-05, "loss": 0.0229, "step": 122390 }, { "epoch": 2.4459454058590784, "grad_norm": 0.13969126343727112, "learning_rate": 1.4354302264204994e-05, "loss": 0.0346, "step": 122400 }, { "epoch": 2.44614523800008, "grad_norm": 0.10150125622749329, "learning_rate": 1.4347056161731038e-05, "loss": 0.0298, "step": 122410 }, { "epoch": 2.4463450701410814, "grad_norm": 0.15816852450370789, "learning_rate": 1.4339811582285318e-05, "loss": 0.0303, "step": 122420 }, { "epoch": 2.446544902282083, "grad_norm": 0.11386743187904358, "learning_rate": 1.4332568526177353e-05, "loss": 0.0224, "step": 122430 }, { "epoch": 2.446744734423085, "grad_norm": 0.13718728721141815, "learning_rate": 1.4325326993716554e-05, "loss": 0.0252, "step": 122440 }, { "epoch": 2.446944566564086, "grad_norm": 0.1132359728217125, "learning_rate": 1.4318086985212232e-05, "loss": 0.0215, "step": 122450 }, { "epoch": 2.447144398705088, "grad_norm": 0.16387425363063812, "learning_rate": 1.4310848500973684e-05, "loss": 0.0272, "step": 122460 }, { "epoch": 2.447344230846089, "grad_norm": 0.12789075076580048, "learning_rate": 1.4303611541310125e-05, "loss": 0.026, "step": 122470 }, { "epoch": 2.447544062987091, "grad_norm": 0.11397959291934967, "learning_rate": 1.4296376106530707e-05, "loss": 0.021, "step": 122480 }, { "epoch": 2.4477438951280925, "grad_norm": 0.12068131566047668, "learning_rate": 1.4289142196944505e-05, "loss": 0.0222, "step": 122490 }, { "epoch": 2.447943727269094, "grad_norm": 0.17431515455245972, "learning_rate": 1.4281909812860556e-05, "loss": 0.0276, "step": 122500 }, { "epoch": 2.4481435594100955, "grad_norm": 0.11751672625541687, "learning_rate": 1.4274678954587794e-05, "loss": 0.0231, "step": 122510 }, { "epoch": 2.4483433915510973, "grad_norm": 0.10454171150922775, "learning_rate": 1.4267449622435125e-05, "loss": 0.0208, "step": 122520 }, { "epoch": 2.4485432236920985, "grad_norm": 0.10121265798807144, "learning_rate": 1.4260221816711362e-05, "loss": 0.0183, "step": 122530 }, { "epoch": 2.4487430558331003, "grad_norm": 0.12099424004554749, "learning_rate": 1.4252995537725277e-05, "loss": 0.0241, "step": 122540 }, { "epoch": 2.448942887974102, "grad_norm": 0.1350029706954956, "learning_rate": 1.4245770785785545e-05, "loss": 0.0268, "step": 122550 }, { "epoch": 2.4491427201151033, "grad_norm": 0.08719179779291153, "learning_rate": 1.4238547561200788e-05, "loss": 0.0231, "step": 122560 }, { "epoch": 2.449342552256105, "grad_norm": 0.09611915796995163, "learning_rate": 1.423132586427961e-05, "loss": 0.0202, "step": 122570 }, { "epoch": 2.4495423843971063, "grad_norm": 0.12699833512306213, "learning_rate": 1.422410569533047e-05, "loss": 0.0215, "step": 122580 }, { "epoch": 2.449742216538108, "grad_norm": 0.1478792130947113, "learning_rate": 1.421688705466181e-05, "loss": 0.0211, "step": 122590 }, { "epoch": 2.4499420486791097, "grad_norm": 0.11676749587059021, "learning_rate": 1.4209669942582e-05, "loss": 0.0245, "step": 122600 }, { "epoch": 2.450141880820111, "grad_norm": 0.12053004652261734, "learning_rate": 1.4202454359399336e-05, "loss": 0.0231, "step": 122610 }, { "epoch": 2.4503417129611127, "grad_norm": 0.1176963523030281, "learning_rate": 1.4195240305422065e-05, "loss": 0.023, "step": 122620 }, { "epoch": 2.4505415451021144, "grad_norm": 0.12411829829216003, "learning_rate": 1.4188027780958352e-05, "loss": 0.0212, "step": 122630 }, { "epoch": 2.4507413772431157, "grad_norm": 0.12189524620771408, "learning_rate": 1.418081678631632e-05, "loss": 0.0251, "step": 122640 }, { "epoch": 2.4509412093841174, "grad_norm": 0.12233779579401016, "learning_rate": 1.4173607321803961e-05, "loss": 0.0298, "step": 122650 }, { "epoch": 2.451141041525119, "grad_norm": 0.12203964591026306, "learning_rate": 1.4166399387729295e-05, "loss": 0.025, "step": 122660 }, { "epoch": 2.4513408736661204, "grad_norm": 0.1008111834526062, "learning_rate": 1.4159192984400237e-05, "loss": 0.0202, "step": 122670 }, { "epoch": 2.451540705807122, "grad_norm": 0.08331547677516937, "learning_rate": 1.4151988112124597e-05, "loss": 0.0259, "step": 122680 }, { "epoch": 2.4517405379481234, "grad_norm": 0.1088421642780304, "learning_rate": 1.4144784771210156e-05, "loss": 0.0284, "step": 122690 }, { "epoch": 2.451940370089125, "grad_norm": 0.10026098787784576, "learning_rate": 1.4137582961964656e-05, "loss": 0.0257, "step": 122700 }, { "epoch": 2.452140202230127, "grad_norm": 0.09985244274139404, "learning_rate": 1.4130382684695742e-05, "loss": 0.0306, "step": 122710 }, { "epoch": 2.452340034371128, "grad_norm": 0.12830474972724915, "learning_rate": 1.4123183939710977e-05, "loss": 0.0327, "step": 122720 }, { "epoch": 2.45253986651213, "grad_norm": 0.06456662714481354, "learning_rate": 1.4115986727317881e-05, "loss": 0.0256, "step": 122730 }, { "epoch": 2.452739698653131, "grad_norm": 0.1676056683063507, "learning_rate": 1.410879104782391e-05, "loss": 0.0213, "step": 122740 }, { "epoch": 2.452939530794133, "grad_norm": 0.21700333058834076, "learning_rate": 1.4101596901536458e-05, "loss": 0.0255, "step": 122750 }, { "epoch": 2.4531393629351346, "grad_norm": 0.13872116804122925, "learning_rate": 1.409440428876283e-05, "loss": 0.0218, "step": 122760 }, { "epoch": 2.4533391950761363, "grad_norm": 0.12829922139644623, "learning_rate": 1.4087213209810297e-05, "loss": 0.0301, "step": 122770 }, { "epoch": 2.4535390272171376, "grad_norm": 0.13462083041667938, "learning_rate": 1.4080023664986041e-05, "loss": 0.021, "step": 122780 }, { "epoch": 2.4537388593581393, "grad_norm": 0.10963567346334457, "learning_rate": 1.4072835654597183e-05, "loss": 0.0254, "step": 122790 }, { "epoch": 2.4539386914991406, "grad_norm": 0.17414061725139618, "learning_rate": 1.4065649178950791e-05, "loss": 0.0267, "step": 122800 }, { "epoch": 2.4541385236401423, "grad_norm": 0.13580207526683807, "learning_rate": 1.4058464238353863e-05, "loss": 0.0327, "step": 122810 }, { "epoch": 2.454338355781144, "grad_norm": 0.1043907180428505, "learning_rate": 1.40512808331133e-05, "loss": 0.0223, "step": 122820 }, { "epoch": 2.4545381879221453, "grad_norm": 0.13778676092624664, "learning_rate": 1.4044098963535967e-05, "loss": 0.0313, "step": 122830 }, { "epoch": 2.454738020063147, "grad_norm": 0.10599873214960098, "learning_rate": 1.4036918629928703e-05, "loss": 0.0316, "step": 122840 }, { "epoch": 2.4549378522041483, "grad_norm": 0.13689911365509033, "learning_rate": 1.4029739832598199e-05, "loss": 0.0256, "step": 122850 }, { "epoch": 2.45513768434515, "grad_norm": 0.10186750441789627, "learning_rate": 1.4022562571851122e-05, "loss": 0.0255, "step": 122860 }, { "epoch": 2.4553375164861517, "grad_norm": 0.12078125029802322, "learning_rate": 1.4015386847994084e-05, "loss": 0.0233, "step": 122870 }, { "epoch": 2.455537348627153, "grad_norm": 0.1417386531829834, "learning_rate": 1.4008212661333608e-05, "loss": 0.0253, "step": 122880 }, { "epoch": 2.4557371807681547, "grad_norm": 0.13638928532600403, "learning_rate": 1.4001040012176175e-05, "loss": 0.0257, "step": 122890 }, { "epoch": 2.4559370129091564, "grad_norm": 0.10486810654401779, "learning_rate": 1.3993868900828178e-05, "loss": 0.0253, "step": 122900 }, { "epoch": 2.4561368450501577, "grad_norm": 0.09754016995429993, "learning_rate": 1.398669932759597e-05, "loss": 0.0349, "step": 122910 }, { "epoch": 2.4563366771911594, "grad_norm": 0.1399550586938858, "learning_rate": 1.397953129278578e-05, "loss": 0.0276, "step": 122920 }, { "epoch": 2.456536509332161, "grad_norm": 0.10462982952594757, "learning_rate": 1.3972364796703857e-05, "loss": 0.0165, "step": 122930 }, { "epoch": 2.4567363414731624, "grad_norm": 0.10526765137910843, "learning_rate": 1.396519983965634e-05, "loss": 0.0221, "step": 122940 }, { "epoch": 2.456936173614164, "grad_norm": 0.14379185438156128, "learning_rate": 1.3958036421949272e-05, "loss": 0.0324, "step": 122950 }, { "epoch": 2.4571360057551654, "grad_norm": 0.16162225604057312, "learning_rate": 1.3950874543888659e-05, "loss": 0.0237, "step": 122960 }, { "epoch": 2.457335837896167, "grad_norm": 0.14315402507781982, "learning_rate": 1.3943714205780495e-05, "loss": 0.0358, "step": 122970 }, { "epoch": 2.457535670037169, "grad_norm": 0.12190882861614227, "learning_rate": 1.3936555407930607e-05, "loss": 0.0281, "step": 122980 }, { "epoch": 2.45773550217817, "grad_norm": 0.1104002594947815, "learning_rate": 1.3929398150644818e-05, "loss": 0.0234, "step": 122990 }, { "epoch": 2.457935334319172, "grad_norm": 0.13771703839302063, "learning_rate": 1.3922242434228877e-05, "loss": 0.0225, "step": 123000 }, { "epoch": 2.4581351664601736, "grad_norm": 0.17190325260162354, "learning_rate": 1.3915088258988467e-05, "loss": 0.0213, "step": 123010 }, { "epoch": 2.458334998601175, "grad_norm": 0.12547776103019714, "learning_rate": 1.390793562522919e-05, "loss": 0.0279, "step": 123020 }, { "epoch": 2.4585348307421766, "grad_norm": 0.12127271294593811, "learning_rate": 1.3900784533256606e-05, "loss": 0.0306, "step": 123030 }, { "epoch": 2.4587346628831783, "grad_norm": 0.12484940886497498, "learning_rate": 1.3893634983376202e-05, "loss": 0.0242, "step": 123040 }, { "epoch": 2.4589344950241796, "grad_norm": 0.11762568354606628, "learning_rate": 1.3886486975893354e-05, "loss": 0.0191, "step": 123050 }, { "epoch": 2.4591343271651813, "grad_norm": 0.10517234355211258, "learning_rate": 1.3879340511113454e-05, "loss": 0.0262, "step": 123060 }, { "epoch": 2.4593341593061826, "grad_norm": 0.12113252282142639, "learning_rate": 1.3872195589341786e-05, "loss": 0.0271, "step": 123070 }, { "epoch": 2.4595339914471843, "grad_norm": 0.09120963513851166, "learning_rate": 1.3865052210883539e-05, "loss": 0.0181, "step": 123080 }, { "epoch": 2.459733823588186, "grad_norm": 0.11540114879608154, "learning_rate": 1.3857910376043864e-05, "loss": 0.0296, "step": 123090 }, { "epoch": 2.4599336557291873, "grad_norm": 0.1267726719379425, "learning_rate": 1.3850770085127873e-05, "loss": 0.0201, "step": 123100 }, { "epoch": 2.460133487870189, "grad_norm": 0.15569114685058594, "learning_rate": 1.3843631338440598e-05, "loss": 0.0235, "step": 123110 }, { "epoch": 2.4603333200111908, "grad_norm": 0.12411289662122726, "learning_rate": 1.3836494136286953e-05, "loss": 0.0249, "step": 123120 }, { "epoch": 2.460533152152192, "grad_norm": 0.25274378061294556, "learning_rate": 1.3829358478971843e-05, "loss": 0.0272, "step": 123130 }, { "epoch": 2.4607329842931938, "grad_norm": 0.15288224816322327, "learning_rate": 1.3822224366800096e-05, "loss": 0.02, "step": 123140 }, { "epoch": 2.4609328164341955, "grad_norm": 0.0857752338051796, "learning_rate": 1.3815091800076457e-05, "loss": 0.0228, "step": 123150 }, { "epoch": 2.4611326485751968, "grad_norm": 0.10180262476205826, "learning_rate": 1.3807960779105627e-05, "loss": 0.0241, "step": 123160 }, { "epoch": 2.4613324807161985, "grad_norm": 0.11698060482740402, "learning_rate": 1.3800831304192225e-05, "loss": 0.02, "step": 123170 }, { "epoch": 2.4615323128571998, "grad_norm": 0.09985685348510742, "learning_rate": 1.3793703375640804e-05, "loss": 0.0237, "step": 123180 }, { "epoch": 2.4617321449982015, "grad_norm": 0.09502702206373215, "learning_rate": 1.3786576993755868e-05, "loss": 0.0326, "step": 123190 }, { "epoch": 2.461931977139203, "grad_norm": 0.1239268109202385, "learning_rate": 1.3779452158841832e-05, "loss": 0.0251, "step": 123200 }, { "epoch": 2.4621318092802045, "grad_norm": 0.14451339840888977, "learning_rate": 1.3772328871203083e-05, "loss": 0.0184, "step": 123210 }, { "epoch": 2.462331641421206, "grad_norm": 0.10678528994321823, "learning_rate": 1.376520713114387e-05, "loss": 0.0314, "step": 123220 }, { "epoch": 2.462531473562208, "grad_norm": 0.13030020892620087, "learning_rate": 1.3758086938968434e-05, "loss": 0.0219, "step": 123230 }, { "epoch": 2.462731305703209, "grad_norm": 0.0962328240275383, "learning_rate": 1.375096829498097e-05, "loss": 0.0207, "step": 123240 }, { "epoch": 2.462931137844211, "grad_norm": 0.08285758644342422, "learning_rate": 1.3743851199485536e-05, "loss": 0.033, "step": 123250 }, { "epoch": 2.4631309699852126, "grad_norm": 0.2122819721698761, "learning_rate": 1.3736735652786176e-05, "loss": 0.0259, "step": 123260 }, { "epoch": 2.463330802126214, "grad_norm": 0.10776542872190475, "learning_rate": 1.3729621655186848e-05, "loss": 0.0191, "step": 123270 }, { "epoch": 2.4635306342672156, "grad_norm": 0.14832989871501923, "learning_rate": 1.3722509206991451e-05, "loss": 0.025, "step": 123280 }, { "epoch": 2.463730466408217, "grad_norm": 0.12008559703826904, "learning_rate": 1.3715398308503818e-05, "loss": 0.0225, "step": 123290 }, { "epoch": 2.4639302985492186, "grad_norm": 0.09778216481208801, "learning_rate": 1.3708288960027716e-05, "loss": 0.0315, "step": 123300 }, { "epoch": 2.4641301306902204, "grad_norm": 0.31981852650642395, "learning_rate": 1.370118116186685e-05, "loss": 0.0283, "step": 123310 }, { "epoch": 2.4643299628312216, "grad_norm": 0.10768692940473557, "learning_rate": 1.3694074914324811e-05, "loss": 0.0252, "step": 123320 }, { "epoch": 2.4645297949722234, "grad_norm": 0.14800523221492767, "learning_rate": 1.3686970217705214e-05, "loss": 0.024, "step": 123330 }, { "epoch": 2.464729627113225, "grad_norm": 0.16019049286842346, "learning_rate": 1.3679867072311559e-05, "loss": 0.0222, "step": 123340 }, { "epoch": 2.4649294592542264, "grad_norm": 0.129558265209198, "learning_rate": 1.3672765478447236e-05, "loss": 0.0226, "step": 123350 }, { "epoch": 2.465129291395228, "grad_norm": 0.11278838664293289, "learning_rate": 1.3665665436415631e-05, "loss": 0.0187, "step": 123360 }, { "epoch": 2.46532912353623, "grad_norm": 0.1009213849902153, "learning_rate": 1.3658566946520074e-05, "loss": 0.0236, "step": 123370 }, { "epoch": 2.465528955677231, "grad_norm": 0.13398770987987518, "learning_rate": 1.365147000906376e-05, "loss": 0.0176, "step": 123380 }, { "epoch": 2.465728787818233, "grad_norm": 0.10001638531684875, "learning_rate": 1.3644374624349876e-05, "loss": 0.0255, "step": 123390 }, { "epoch": 2.465928619959234, "grad_norm": 0.14187034964561462, "learning_rate": 1.3637280792681517e-05, "loss": 0.0237, "step": 123400 }, { "epoch": 2.466128452100236, "grad_norm": 0.1066102460026741, "learning_rate": 1.363018851436173e-05, "loss": 0.0242, "step": 123410 }, { "epoch": 2.4663282842412375, "grad_norm": 0.1192694678902626, "learning_rate": 1.3623097789693467e-05, "loss": 0.023, "step": 123420 }, { "epoch": 2.466528116382239, "grad_norm": 0.11632467806339264, "learning_rate": 1.3616008618979648e-05, "loss": 0.0236, "step": 123430 }, { "epoch": 2.4667279485232405, "grad_norm": 0.11013295501470566, "learning_rate": 1.3608921002523111e-05, "loss": 0.03, "step": 123440 }, { "epoch": 2.466927780664242, "grad_norm": 0.13170230388641357, "learning_rate": 1.3601834940626585e-05, "loss": 0.0181, "step": 123450 }, { "epoch": 2.4671276128052435, "grad_norm": 0.10694590955972672, "learning_rate": 1.3594750433592824e-05, "loss": 0.0246, "step": 123460 }, { "epoch": 2.4673274449462452, "grad_norm": 0.18978406488895416, "learning_rate": 1.3587667481724442e-05, "loss": 0.0171, "step": 123470 }, { "epoch": 2.467527277087247, "grad_norm": 0.16160836815834045, "learning_rate": 1.3580586085324031e-05, "loss": 0.0233, "step": 123480 }, { "epoch": 2.4677271092282482, "grad_norm": 0.1357780545949936, "learning_rate": 1.3573506244694062e-05, "loss": 0.0259, "step": 123490 }, { "epoch": 2.46792694136925, "grad_norm": 0.09793172031641006, "learning_rate": 1.3566427960136974e-05, "loss": 0.0279, "step": 123500 }, { "epoch": 2.4681267735102512, "grad_norm": 0.12158111482858658, "learning_rate": 1.3559351231955187e-05, "loss": 0.0246, "step": 123510 }, { "epoch": 2.468326605651253, "grad_norm": 0.1368943303823471, "learning_rate": 1.3552276060450947e-05, "loss": 0.0241, "step": 123520 }, { "epoch": 2.4685264377922547, "grad_norm": 0.11991389095783234, "learning_rate": 1.3545202445926525e-05, "loss": 0.0187, "step": 123530 }, { "epoch": 2.468726269933256, "grad_norm": 0.14862824976444244, "learning_rate": 1.3538130388684084e-05, "loss": 0.0298, "step": 123540 }, { "epoch": 2.4689261020742577, "grad_norm": 0.11625905334949493, "learning_rate": 1.3531059889025727e-05, "loss": 0.0286, "step": 123550 }, { "epoch": 2.469125934215259, "grad_norm": 0.11560095101594925, "learning_rate": 1.3523990947253502e-05, "loss": 0.0247, "step": 123560 }, { "epoch": 2.4693257663562607, "grad_norm": 0.1732342392206192, "learning_rate": 1.351692356366937e-05, "loss": 0.0304, "step": 123570 }, { "epoch": 2.4695255984972624, "grad_norm": 0.12603861093521118, "learning_rate": 1.3509857738575254e-05, "loss": 0.0333, "step": 123580 }, { "epoch": 2.469725430638264, "grad_norm": 0.09311246871948242, "learning_rate": 1.3502793472272957e-05, "loss": 0.0232, "step": 123590 }, { "epoch": 2.4699252627792654, "grad_norm": 0.11762578040361404, "learning_rate": 1.3495730765064285e-05, "loss": 0.0348, "step": 123600 }, { "epoch": 2.470125094920267, "grad_norm": 0.22537054121494293, "learning_rate": 1.3488669617250954e-05, "loss": 0.0153, "step": 123610 }, { "epoch": 2.4703249270612684, "grad_norm": 0.09277432411909103, "learning_rate": 1.3481610029134568e-05, "loss": 0.0163, "step": 123620 }, { "epoch": 2.47052475920227, "grad_norm": 0.18054857850074768, "learning_rate": 1.3474552001016699e-05, "loss": 0.0312, "step": 123630 }, { "epoch": 2.470724591343272, "grad_norm": 0.19538167119026184, "learning_rate": 1.3467495533198898e-05, "loss": 0.0214, "step": 123640 }, { "epoch": 2.470924423484273, "grad_norm": 0.10451352596282959, "learning_rate": 1.3460440625982563e-05, "loss": 0.0214, "step": 123650 }, { "epoch": 2.471124255625275, "grad_norm": 0.12474735081195831, "learning_rate": 1.3453387279669078e-05, "loss": 0.0284, "step": 123660 }, { "epoch": 2.471324087766276, "grad_norm": 0.09365280717611313, "learning_rate": 1.3446335494559747e-05, "loss": 0.0196, "step": 123670 }, { "epoch": 2.471523919907278, "grad_norm": 0.11747869104146957, "learning_rate": 1.3439285270955815e-05, "loss": 0.0191, "step": 123680 }, { "epoch": 2.4717237520482795, "grad_norm": 0.1144038736820221, "learning_rate": 1.3432236609158455e-05, "loss": 0.0217, "step": 123690 }, { "epoch": 2.471923584189281, "grad_norm": 0.11689987033605576, "learning_rate": 1.3425189509468766e-05, "loss": 0.0277, "step": 123700 }, { "epoch": 2.4721234163302825, "grad_norm": 0.1325823962688446, "learning_rate": 1.3418143972187809e-05, "loss": 0.0242, "step": 123710 }, { "epoch": 2.4723232484712843, "grad_norm": 0.1004568487405777, "learning_rate": 1.3411099997616505e-05, "loss": 0.0205, "step": 123720 }, { "epoch": 2.4725230806122855, "grad_norm": 0.11360307037830353, "learning_rate": 1.3404057586055813e-05, "loss": 0.0283, "step": 123730 }, { "epoch": 2.4727229127532873, "grad_norm": 0.12394307553768158, "learning_rate": 1.339701673780656e-05, "loss": 0.0221, "step": 123740 }, { "epoch": 2.472922744894289, "grad_norm": 0.1422327607870102, "learning_rate": 1.33899774531695e-05, "loss": 0.0281, "step": 123750 }, { "epoch": 2.4731225770352903, "grad_norm": 0.10739678889513016, "learning_rate": 1.338293973244535e-05, "loss": 0.0209, "step": 123760 }, { "epoch": 2.473322409176292, "grad_norm": 0.09025643020868301, "learning_rate": 1.337590357593474e-05, "loss": 0.0285, "step": 123770 }, { "epoch": 2.4735222413172933, "grad_norm": 0.13708744943141937, "learning_rate": 1.3368868983938255e-05, "loss": 0.0261, "step": 123780 }, { "epoch": 2.473722073458295, "grad_norm": 0.11117470264434814, "learning_rate": 1.3361835956756391e-05, "loss": 0.0204, "step": 123790 }, { "epoch": 2.4739219055992967, "grad_norm": 0.12935318052768707, "learning_rate": 1.3354804494689588e-05, "loss": 0.036, "step": 123800 }, { "epoch": 2.474121737740298, "grad_norm": 0.17376196384429932, "learning_rate": 1.334777459803822e-05, "loss": 0.0262, "step": 123810 }, { "epoch": 2.4743215698812997, "grad_norm": 0.12494908273220062, "learning_rate": 1.3340746267102589e-05, "loss": 0.0268, "step": 123820 }, { "epoch": 2.4745214020223014, "grad_norm": 0.1194809079170227, "learning_rate": 1.3333719502182928e-05, "loss": 0.0283, "step": 123830 }, { "epoch": 2.4747212341633027, "grad_norm": 0.11947955191135406, "learning_rate": 1.3326694303579428e-05, "loss": 0.0259, "step": 123840 }, { "epoch": 2.4749210663043044, "grad_norm": 0.11907808482646942, "learning_rate": 1.3319670671592166e-05, "loss": 0.0201, "step": 123850 }, { "epoch": 2.475120898445306, "grad_norm": 0.12837642431259155, "learning_rate": 1.3312648606521172e-05, "loss": 0.0209, "step": 123860 }, { "epoch": 2.4753207305863074, "grad_norm": 0.1405172049999237, "learning_rate": 1.3305628108666446e-05, "loss": 0.0191, "step": 123870 }, { "epoch": 2.475520562727309, "grad_norm": 0.11490500718355179, "learning_rate": 1.3298609178327892e-05, "loss": 0.021, "step": 123880 }, { "epoch": 2.4757203948683104, "grad_norm": 0.09446194022893906, "learning_rate": 1.3291591815805316e-05, "loss": 0.0193, "step": 123890 }, { "epoch": 2.475920227009312, "grad_norm": 0.1388867199420929, "learning_rate": 1.3284576021398493e-05, "loss": 0.0326, "step": 123900 }, { "epoch": 2.476120059150314, "grad_norm": 0.133046954870224, "learning_rate": 1.3277561795407156e-05, "loss": 0.0184, "step": 123910 }, { "epoch": 2.476319891291315, "grad_norm": 0.07868187874555588, "learning_rate": 1.3270549138130906e-05, "loss": 0.0259, "step": 123920 }, { "epoch": 2.476519723432317, "grad_norm": 0.11788704246282578, "learning_rate": 1.3263538049869322e-05, "loss": 0.0286, "step": 123930 }, { "epoch": 2.4767195555733186, "grad_norm": 0.1410350501537323, "learning_rate": 1.3256528530921907e-05, "loss": 0.0254, "step": 123940 }, { "epoch": 2.47691938771432, "grad_norm": 0.11679708957672119, "learning_rate": 1.3249520581588092e-05, "loss": 0.02, "step": 123950 }, { "epoch": 2.4771192198553216, "grad_norm": 0.20725418627262115, "learning_rate": 1.324251420216725e-05, "loss": 0.0332, "step": 123960 }, { "epoch": 2.4773190519963233, "grad_norm": 0.11612287163734436, "learning_rate": 1.3235509392958668e-05, "loss": 0.0238, "step": 123970 }, { "epoch": 2.4775188841373246, "grad_norm": 0.1280442476272583, "learning_rate": 1.3228506154261605e-05, "loss": 0.0268, "step": 123980 }, { "epoch": 2.4777187162783263, "grad_norm": 0.08680606633424759, "learning_rate": 1.3221504486375175e-05, "loss": 0.0213, "step": 123990 }, { "epoch": 2.4779185484193276, "grad_norm": 0.11104197800159454, "learning_rate": 1.3214504389598531e-05, "loss": 0.0257, "step": 124000 }, { "epoch": 2.4781183805603293, "grad_norm": 0.11527147889137268, "learning_rate": 1.320750586423069e-05, "loss": 0.0286, "step": 124010 }, { "epoch": 2.478318212701331, "grad_norm": 0.10922842472791672, "learning_rate": 1.3200508910570597e-05, "loss": 0.0211, "step": 124020 }, { "epoch": 2.4785180448423323, "grad_norm": 0.10574501752853394, "learning_rate": 1.3193513528917156e-05, "loss": 0.0247, "step": 124030 }, { "epoch": 2.478717876983334, "grad_norm": 0.10663339495658875, "learning_rate": 1.3186519719569207e-05, "loss": 0.0319, "step": 124040 }, { "epoch": 2.4789177091243357, "grad_norm": 0.12495321780443192, "learning_rate": 1.3179527482825504e-05, "loss": 0.0239, "step": 124050 }, { "epoch": 2.479117541265337, "grad_norm": 0.0827980563044548, "learning_rate": 1.3172536818984749e-05, "loss": 0.0265, "step": 124060 }, { "epoch": 2.4793173734063387, "grad_norm": 0.10893457382917404, "learning_rate": 1.3165547728345562e-05, "loss": 0.0252, "step": 124070 }, { "epoch": 2.4795172055473405, "grad_norm": 0.11151804029941559, "learning_rate": 1.315856021120651e-05, "loss": 0.0265, "step": 124080 }, { "epoch": 2.4797170376883417, "grad_norm": 0.15361283719539642, "learning_rate": 1.315157426786609e-05, "loss": 0.024, "step": 124090 }, { "epoch": 2.4799168698293435, "grad_norm": 0.13483715057373047, "learning_rate": 1.3144589898622722e-05, "loss": 0.0194, "step": 124100 }, { "epoch": 2.4801167019703447, "grad_norm": 0.11463790386915207, "learning_rate": 1.3137607103774786e-05, "loss": 0.0218, "step": 124110 }, { "epoch": 2.4803165341113464, "grad_norm": 0.12017691880464554, "learning_rate": 1.3130625883620535e-05, "loss": 0.0364, "step": 124120 }, { "epoch": 2.480516366252348, "grad_norm": 0.11385758221149445, "learning_rate": 1.3123646238458204e-05, "loss": 0.0216, "step": 124130 }, { "epoch": 2.4807161983933494, "grad_norm": 0.14244326949119568, "learning_rate": 1.3116668168585987e-05, "loss": 0.0164, "step": 124140 }, { "epoch": 2.480916030534351, "grad_norm": 0.10602811723947525, "learning_rate": 1.3109691674301944e-05, "loss": 0.0226, "step": 124150 }, { "epoch": 2.481115862675353, "grad_norm": 0.09348193556070328, "learning_rate": 1.3102716755904092e-05, "loss": 0.0351, "step": 124160 }, { "epoch": 2.481315694816354, "grad_norm": 0.1675921529531479, "learning_rate": 1.3095743413690404e-05, "loss": 0.0292, "step": 124170 }, { "epoch": 2.481515526957356, "grad_norm": 0.10567399859428406, "learning_rate": 1.308877164795876e-05, "loss": 0.0184, "step": 124180 }, { "epoch": 2.4817153590983576, "grad_norm": 0.11366177350282669, "learning_rate": 1.3081801459006986e-05, "loss": 0.0249, "step": 124190 }, { "epoch": 2.481915191239359, "grad_norm": 0.09442390501499176, "learning_rate": 1.307483284713283e-05, "loss": 0.0221, "step": 124200 }, { "epoch": 2.4821150233803606, "grad_norm": 0.1173495203256607, "learning_rate": 1.3067865812633996e-05, "loss": 0.0287, "step": 124210 }, { "epoch": 2.482314855521362, "grad_norm": 0.13556864857673645, "learning_rate": 1.3060900355808059e-05, "loss": 0.0283, "step": 124220 }, { "epoch": 2.4825146876623636, "grad_norm": 0.11101926118135452, "learning_rate": 1.305393647695261e-05, "loss": 0.0207, "step": 124230 }, { "epoch": 2.4827145198033653, "grad_norm": 0.1236511841416359, "learning_rate": 1.3046974176365146e-05, "loss": 0.036, "step": 124240 }, { "epoch": 2.4829143519443666, "grad_norm": 0.1293877512216568, "learning_rate": 1.3040013454343041e-05, "loss": 0.0219, "step": 124250 }, { "epoch": 2.4831141840853683, "grad_norm": 0.12571756541728973, "learning_rate": 1.3033054311183645e-05, "loss": 0.0197, "step": 124260 }, { "epoch": 2.4833140162263696, "grad_norm": 0.12351533770561218, "learning_rate": 1.302609674718428e-05, "loss": 0.0171, "step": 124270 }, { "epoch": 2.4835138483673713, "grad_norm": 0.13569316267967224, "learning_rate": 1.3019140762642152e-05, "loss": 0.0234, "step": 124280 }, { "epoch": 2.483713680508373, "grad_norm": 0.13283798098564148, "learning_rate": 1.3012186357854378e-05, "loss": 0.0321, "step": 124290 }, { "epoch": 2.4839135126493748, "grad_norm": 0.1420985907316208, "learning_rate": 1.3005233533118038e-05, "loss": 0.0359, "step": 124300 }, { "epoch": 2.484113344790376, "grad_norm": 0.14239870011806488, "learning_rate": 1.299828228873019e-05, "loss": 0.025, "step": 124310 }, { "epoch": 2.4843131769313778, "grad_norm": 0.12185141444206238, "learning_rate": 1.2991332624987734e-05, "loss": 0.0232, "step": 124320 }, { "epoch": 2.484513009072379, "grad_norm": 0.14221088588237762, "learning_rate": 1.2984384542187562e-05, "loss": 0.0262, "step": 124330 }, { "epoch": 2.4847128412133808, "grad_norm": 0.12587210536003113, "learning_rate": 1.2977438040626483e-05, "loss": 0.0233, "step": 124340 }, { "epoch": 2.4849126733543825, "grad_norm": 0.14454923570156097, "learning_rate": 1.2970493120601235e-05, "loss": 0.0226, "step": 124350 }, { "epoch": 2.4851125054953838, "grad_norm": 0.12005055695772171, "learning_rate": 1.2963549782408502e-05, "loss": 0.0228, "step": 124360 }, { "epoch": 2.4853123376363855, "grad_norm": 0.10859926789999008, "learning_rate": 1.295660802634488e-05, "loss": 0.0258, "step": 124370 }, { "epoch": 2.4855121697773868, "grad_norm": 0.13903303444385529, "learning_rate": 1.2949667852706927e-05, "loss": 0.0224, "step": 124380 }, { "epoch": 2.4857120019183885, "grad_norm": 0.08473935723304749, "learning_rate": 1.294272926179107e-05, "loss": 0.0252, "step": 124390 }, { "epoch": 2.48591183405939, "grad_norm": 0.14659151434898376, "learning_rate": 1.2935792253893757e-05, "loss": 0.0243, "step": 124400 }, { "epoch": 2.4861116662003915, "grad_norm": 0.10366933047771454, "learning_rate": 1.2928856829311326e-05, "loss": 0.028, "step": 124410 }, { "epoch": 2.486311498341393, "grad_norm": 0.12852182984352112, "learning_rate": 1.2921922988340012e-05, "loss": 0.0295, "step": 124420 }, { "epoch": 2.486511330482395, "grad_norm": 0.13484764099121094, "learning_rate": 1.2914990731276034e-05, "loss": 0.0244, "step": 124430 }, { "epoch": 2.486711162623396, "grad_norm": 0.1295941323041916, "learning_rate": 1.2908060058415527e-05, "loss": 0.0255, "step": 124440 }, { "epoch": 2.486910994764398, "grad_norm": 0.10998301208019257, "learning_rate": 1.2901130970054554e-05, "loss": 0.0237, "step": 124450 }, { "epoch": 2.4871108269053996, "grad_norm": 0.12591995298862457, "learning_rate": 1.2894203466489107e-05, "loss": 0.0342, "step": 124460 }, { "epoch": 2.487310659046401, "grad_norm": 0.17024242877960205, "learning_rate": 1.2887277548015131e-05, "loss": 0.0274, "step": 124470 }, { "epoch": 2.4875104911874026, "grad_norm": 0.3001922070980072, "learning_rate": 1.2880353214928475e-05, "loss": 0.0318, "step": 124480 }, { "epoch": 2.487710323328404, "grad_norm": 0.11213633418083191, "learning_rate": 1.2873430467524938e-05, "loss": 0.0218, "step": 124490 }, { "epoch": 2.4879101554694056, "grad_norm": 0.12022354453802109, "learning_rate": 1.2866509306100256e-05, "loss": 0.0236, "step": 124500 }, { "epoch": 2.4881099876104074, "grad_norm": 0.14686737954616547, "learning_rate": 1.2859589730950083e-05, "loss": 0.0279, "step": 124510 }, { "epoch": 2.4883098197514086, "grad_norm": 0.13616788387298584, "learning_rate": 1.2852671742369998e-05, "loss": 0.0269, "step": 124520 }, { "epoch": 2.4885096518924104, "grad_norm": 0.11957193166017532, "learning_rate": 1.2845755340655525e-05, "loss": 0.0217, "step": 124530 }, { "epoch": 2.488709484033412, "grad_norm": 0.10519206523895264, "learning_rate": 1.2838840526102153e-05, "loss": 0.0272, "step": 124540 }, { "epoch": 2.4889093161744134, "grad_norm": 0.08736921846866608, "learning_rate": 1.2831927299005231e-05, "loss": 0.0263, "step": 124550 }, { "epoch": 2.489109148315415, "grad_norm": 0.11649464070796967, "learning_rate": 1.2825015659660094e-05, "loss": 0.0205, "step": 124560 }, { "epoch": 2.489308980456417, "grad_norm": 0.0941728726029396, "learning_rate": 1.2818105608362003e-05, "loss": 0.0255, "step": 124570 }, { "epoch": 2.489508812597418, "grad_norm": 0.09781517088413239, "learning_rate": 1.2811197145406128e-05, "loss": 0.0239, "step": 124580 }, { "epoch": 2.48970864473842, "grad_norm": 0.09293213486671448, "learning_rate": 1.2804290271087594e-05, "loss": 0.0197, "step": 124590 }, { "epoch": 2.489908476879421, "grad_norm": 0.09432687610387802, "learning_rate": 1.279738498570145e-05, "loss": 0.0229, "step": 124600 }, { "epoch": 2.490108309020423, "grad_norm": 0.14946182072162628, "learning_rate": 1.2790481289542689e-05, "loss": 0.0278, "step": 124610 }, { "epoch": 2.4903081411614245, "grad_norm": 0.11029944568872452, "learning_rate": 1.2783579182906185e-05, "loss": 0.0199, "step": 124620 }, { "epoch": 2.490507973302426, "grad_norm": 0.16281858086585999, "learning_rate": 1.2776678666086822e-05, "loss": 0.0217, "step": 124630 }, { "epoch": 2.4907078054434275, "grad_norm": 0.09156440198421478, "learning_rate": 1.2769779739379378e-05, "loss": 0.0186, "step": 124640 }, { "epoch": 2.4909076375844292, "grad_norm": 0.13702227175235748, "learning_rate": 1.2762882403078541e-05, "loss": 0.0306, "step": 124650 }, { "epoch": 2.4911074697254305, "grad_norm": 0.11350630223751068, "learning_rate": 1.2755986657478942e-05, "loss": 0.023, "step": 124660 }, { "epoch": 2.4913073018664322, "grad_norm": 0.12303245067596436, "learning_rate": 1.2749092502875188e-05, "loss": 0.0244, "step": 124670 }, { "epoch": 2.491507134007434, "grad_norm": 0.09908092021942139, "learning_rate": 1.2742199939561783e-05, "loss": 0.0253, "step": 124680 }, { "epoch": 2.4917069661484352, "grad_norm": 0.12073414772748947, "learning_rate": 1.2735308967833142e-05, "loss": 0.0237, "step": 124690 }, { "epoch": 2.491906798289437, "grad_norm": 0.11702427268028259, "learning_rate": 1.2728419587983642e-05, "loss": 0.0327, "step": 124700 }, { "epoch": 2.4921066304304382, "grad_norm": 0.15318123996257782, "learning_rate": 1.2721531800307584e-05, "loss": 0.0186, "step": 124710 }, { "epoch": 2.49230646257144, "grad_norm": 0.10098154842853546, "learning_rate": 1.2714645605099207e-05, "loss": 0.0258, "step": 124720 }, { "epoch": 2.4925062947124417, "grad_norm": 0.09397730976343155, "learning_rate": 1.2707761002652679e-05, "loss": 0.0276, "step": 124730 }, { "epoch": 2.492706126853443, "grad_norm": 0.10386770218610764, "learning_rate": 1.2700877993262089e-05, "loss": 0.0277, "step": 124740 }, { "epoch": 2.4929059589944447, "grad_norm": 0.1414862722158432, "learning_rate": 1.2693996577221467e-05, "loss": 0.0286, "step": 124750 }, { "epoch": 2.4931057911354464, "grad_norm": 0.13263937830924988, "learning_rate": 1.2687116754824784e-05, "loss": 0.0356, "step": 124760 }, { "epoch": 2.4933056232764477, "grad_norm": 0.11566653102636337, "learning_rate": 1.2680238526365923e-05, "loss": 0.0238, "step": 124770 }, { "epoch": 2.4935054554174494, "grad_norm": 0.11432967334985733, "learning_rate": 1.267336189213873e-05, "loss": 0.0297, "step": 124780 }, { "epoch": 2.493705287558451, "grad_norm": 0.12126289308071136, "learning_rate": 1.2666486852436933e-05, "loss": 0.0214, "step": 124790 }, { "epoch": 2.4939051196994524, "grad_norm": 0.10393568128347397, "learning_rate": 1.2659613407554216e-05, "loss": 0.0273, "step": 124800 }, { "epoch": 2.494104951840454, "grad_norm": 0.1105705201625824, "learning_rate": 1.2652741557784247e-05, "loss": 0.028, "step": 124810 }, { "epoch": 2.4943047839814554, "grad_norm": 0.12319561839103699, "learning_rate": 1.2645871303420537e-05, "loss": 0.0224, "step": 124820 }, { "epoch": 2.494504616122457, "grad_norm": 0.10549981147050858, "learning_rate": 1.2639002644756581e-05, "loss": 0.0192, "step": 124830 }, { "epoch": 2.494704448263459, "grad_norm": 0.13332529366016388, "learning_rate": 1.2632135582085803e-05, "loss": 0.02, "step": 124840 }, { "epoch": 2.49490428040446, "grad_norm": 0.12566153705120087, "learning_rate": 1.2625270115701537e-05, "loss": 0.0253, "step": 124850 }, { "epoch": 2.495104112545462, "grad_norm": 0.12184751033782959, "learning_rate": 1.261840624589708e-05, "loss": 0.0347, "step": 124860 }, { "epoch": 2.4953039446864635, "grad_norm": 0.10935687273740768, "learning_rate": 1.2611543972965634e-05, "loss": 0.0207, "step": 124870 }, { "epoch": 2.495503776827465, "grad_norm": 0.16456352174282074, "learning_rate": 1.260468329720036e-05, "loss": 0.0206, "step": 124880 }, { "epoch": 2.4957036089684665, "grad_norm": 0.13433785736560822, "learning_rate": 1.2597824218894283e-05, "loss": 0.0249, "step": 124890 }, { "epoch": 2.4959034411094683, "grad_norm": 0.0903177484869957, "learning_rate": 1.2590966738340465e-05, "loss": 0.0249, "step": 124900 }, { "epoch": 2.4961032732504695, "grad_norm": 0.09664768725633621, "learning_rate": 1.2584110855831838e-05, "loss": 0.022, "step": 124910 }, { "epoch": 2.4963031053914713, "grad_norm": 0.13628971576690674, "learning_rate": 1.257725657166125e-05, "loss": 0.0206, "step": 124920 }, { "epoch": 2.4965029375324725, "grad_norm": 0.11818882822990417, "learning_rate": 1.2570403886121496e-05, "loss": 0.0226, "step": 124930 }, { "epoch": 2.4967027696734743, "grad_norm": 0.11284088343381882, "learning_rate": 1.2563552799505358e-05, "loss": 0.0208, "step": 124940 }, { "epoch": 2.496902601814476, "grad_norm": 0.12098640203475952, "learning_rate": 1.255670331210545e-05, "loss": 0.0214, "step": 124950 }, { "epoch": 2.4971024339554773, "grad_norm": 0.11547297239303589, "learning_rate": 1.2549855424214397e-05, "loss": 0.0263, "step": 124960 }, { "epoch": 2.497302266096479, "grad_norm": 0.12441453337669373, "learning_rate": 1.2543009136124723e-05, "loss": 0.0205, "step": 124970 }, { "epoch": 2.4975020982374807, "grad_norm": 0.13776230812072754, "learning_rate": 1.2536164448128885e-05, "loss": 0.0198, "step": 124980 }, { "epoch": 2.497701930378482, "grad_norm": 0.1276487410068512, "learning_rate": 1.252932136051928e-05, "loss": 0.034, "step": 124990 }, { "epoch": 2.4979017625194837, "grad_norm": 0.13810482621192932, "learning_rate": 1.2522479873588233e-05, "loss": 0.0248, "step": 125000 }, { "epoch": 2.4981015946604854, "grad_norm": 0.10075819492340088, "learning_rate": 1.2515639987628008e-05, "loss": 0.0263, "step": 125010 }, { "epoch": 2.4983014268014867, "grad_norm": 0.11937736719846725, "learning_rate": 1.2508801702930756e-05, "loss": 0.0266, "step": 125020 }, { "epoch": 2.4985012589424884, "grad_norm": 0.09291817247867584, "learning_rate": 1.250196501978863e-05, "loss": 0.023, "step": 125030 }, { "epoch": 2.4987010910834897, "grad_norm": 0.09129543602466583, "learning_rate": 1.2495129938493694e-05, "loss": 0.0195, "step": 125040 }, { "epoch": 2.4989009232244914, "grad_norm": 0.1518157720565796, "learning_rate": 1.248829645933789e-05, "loss": 0.0269, "step": 125050 }, { "epoch": 2.499100755365493, "grad_norm": 0.11427683383226395, "learning_rate": 1.248146458261315e-05, "loss": 0.026, "step": 125060 }, { "epoch": 2.4993005875064944, "grad_norm": 0.13639333844184875, "learning_rate": 1.2474634308611305e-05, "loss": 0.0231, "step": 125070 }, { "epoch": 2.499500419647496, "grad_norm": 0.12158206850290298, "learning_rate": 1.246780563762418e-05, "loss": 0.0288, "step": 125080 }, { "epoch": 2.4997002517884974, "grad_norm": 0.15026773512363434, "learning_rate": 1.2460978569943426e-05, "loss": 0.0238, "step": 125090 }, { "epoch": 2.499900083929499, "grad_norm": 0.1306799352169037, "learning_rate": 1.245415310586071e-05, "loss": 0.027, "step": 125100 }, { "epoch": 2.500099916070501, "grad_norm": 0.18083740770816803, "learning_rate": 1.2447329245667605e-05, "loss": 0.0209, "step": 125110 }, { "epoch": 2.5002997482115026, "grad_norm": 0.11963378638029099, "learning_rate": 1.2440506989655603e-05, "loss": 0.0273, "step": 125120 }, { "epoch": 2.500499580352504, "grad_norm": 0.12479925900697708, "learning_rate": 1.2433686338116141e-05, "loss": 0.0311, "step": 125130 }, { "epoch": 2.5006994124935056, "grad_norm": 0.11918462067842484, "learning_rate": 1.2426867291340593e-05, "loss": 0.0315, "step": 125140 }, { "epoch": 2.500899244634507, "grad_norm": 0.10136964917182922, "learning_rate": 1.2420049849620264e-05, "loss": 0.0252, "step": 125150 }, { "epoch": 2.5010990767755086, "grad_norm": 0.14151979982852936, "learning_rate": 1.2413234013246333e-05, "loss": 0.0235, "step": 125160 }, { "epoch": 2.5012989089165103, "grad_norm": 0.122312992811203, "learning_rate": 1.2406419782510015e-05, "loss": 0.026, "step": 125170 }, { "epoch": 2.5014987410575116, "grad_norm": 0.10032650083303452, "learning_rate": 1.2399607157702398e-05, "loss": 0.0241, "step": 125180 }, { "epoch": 2.5016985731985133, "grad_norm": 0.10783614963293076, "learning_rate": 1.239279613911447e-05, "loss": 0.0205, "step": 125190 }, { "epoch": 2.5018984053395146, "grad_norm": 0.16600291430950165, "learning_rate": 1.2385986727037196e-05, "loss": 0.0204, "step": 125200 }, { "epoch": 2.5020982374805163, "grad_norm": 0.14933806657791138, "learning_rate": 1.2379178921761497e-05, "loss": 0.0246, "step": 125210 }, { "epoch": 2.502298069621518, "grad_norm": 0.1531629115343094, "learning_rate": 1.2372372723578146e-05, "loss": 0.0343, "step": 125220 }, { "epoch": 2.5024979017625197, "grad_norm": 0.128752663731575, "learning_rate": 1.236556813277791e-05, "loss": 0.0173, "step": 125230 }, { "epoch": 2.502697733903521, "grad_norm": 0.11186206340789795, "learning_rate": 1.2358765149651464e-05, "loss": 0.0237, "step": 125240 }, { "epoch": 2.5028975660445227, "grad_norm": 0.10750806331634521, "learning_rate": 1.2351963774489422e-05, "loss": 0.0267, "step": 125250 }, { "epoch": 2.503097398185524, "grad_norm": 0.15346406400203705, "learning_rate": 1.2345164007582332e-05, "loss": 0.0275, "step": 125260 }, { "epoch": 2.5032972303265257, "grad_norm": 0.08690856397151947, "learning_rate": 1.2338365849220657e-05, "loss": 0.0226, "step": 125270 }, { "epoch": 2.5034970624675275, "grad_norm": 0.10606179386377335, "learning_rate": 1.2331569299694818e-05, "loss": 0.0247, "step": 125280 }, { "epoch": 2.5036968946085287, "grad_norm": 0.1899523138999939, "learning_rate": 1.2324774359295116e-05, "loss": 0.0336, "step": 125290 }, { "epoch": 2.5038967267495305, "grad_norm": 0.10508762300014496, "learning_rate": 1.2317981028311854e-05, "loss": 0.027, "step": 125300 }, { "epoch": 2.5040965588905317, "grad_norm": 0.13285325467586517, "learning_rate": 1.2311189307035232e-05, "loss": 0.0238, "step": 125310 }, { "epoch": 2.5042963910315335, "grad_norm": 0.09183113276958466, "learning_rate": 1.2304399195755351e-05, "loss": 0.0249, "step": 125320 }, { "epoch": 2.504496223172535, "grad_norm": 0.1623287945985794, "learning_rate": 1.229761069476229e-05, "loss": 0.0325, "step": 125330 }, { "epoch": 2.504696055313537, "grad_norm": 0.09626980870962143, "learning_rate": 1.2290823804346036e-05, "loss": 0.0209, "step": 125340 }, { "epoch": 2.504895887454538, "grad_norm": 0.1343182921409607, "learning_rate": 1.2284038524796515e-05, "loss": 0.0275, "step": 125350 }, { "epoch": 2.50509571959554, "grad_norm": 0.19424307346343994, "learning_rate": 1.227725485640358e-05, "loss": 0.0311, "step": 125360 }, { "epoch": 2.505295551736541, "grad_norm": 0.13718029856681824, "learning_rate": 1.2270472799457021e-05, "loss": 0.0223, "step": 125370 }, { "epoch": 2.505495383877543, "grad_norm": 0.1377294510602951, "learning_rate": 1.2263692354246559e-05, "loss": 0.0321, "step": 125380 }, { "epoch": 2.5056952160185446, "grad_norm": 0.10063055902719498, "learning_rate": 1.225691352106183e-05, "loss": 0.022, "step": 125390 }, { "epoch": 2.505895048159546, "grad_norm": 0.13044853508472443, "learning_rate": 1.2250136300192417e-05, "loss": 0.0206, "step": 125400 }, { "epoch": 2.5060948803005476, "grad_norm": 0.1164693832397461, "learning_rate": 1.2243360691927846e-05, "loss": 0.0222, "step": 125410 }, { "epoch": 2.506294712441549, "grad_norm": 0.11528037488460541, "learning_rate": 1.223658669655754e-05, "loss": 0.024, "step": 125420 }, { "epoch": 2.5064945445825506, "grad_norm": 0.12083367258310318, "learning_rate": 1.2229814314370852e-05, "loss": 0.0239, "step": 125430 }, { "epoch": 2.5066943767235523, "grad_norm": 0.09687919914722443, "learning_rate": 1.2223043545657143e-05, "loss": 0.0216, "step": 125440 }, { "epoch": 2.506894208864554, "grad_norm": 0.1251201033592224, "learning_rate": 1.2216274390705601e-05, "loss": 0.0303, "step": 125450 }, { "epoch": 2.5070940410055553, "grad_norm": 0.14019176363945007, "learning_rate": 1.2209506849805408e-05, "loss": 0.0298, "step": 125460 }, { "epoch": 2.507293873146557, "grad_norm": 0.09970906376838684, "learning_rate": 1.2202740923245642e-05, "loss": 0.0309, "step": 125470 }, { "epoch": 2.5074937052875583, "grad_norm": 0.13619649410247803, "learning_rate": 1.219597661131538e-05, "loss": 0.025, "step": 125480 }, { "epoch": 2.50769353742856, "grad_norm": 0.13415947556495667, "learning_rate": 1.2189213914303532e-05, "loss": 0.0208, "step": 125490 }, { "epoch": 2.5078933695695618, "grad_norm": 0.13990706205368042, "learning_rate": 1.2182452832499003e-05, "loss": 0.022, "step": 125500 }, { "epoch": 2.508093201710563, "grad_norm": 0.09179817885160446, "learning_rate": 1.217569336619062e-05, "loss": 0.0169, "step": 125510 }, { "epoch": 2.5082930338515648, "grad_norm": 0.09621214121580124, "learning_rate": 1.216893551566713e-05, "loss": 0.0196, "step": 125520 }, { "epoch": 2.508492865992566, "grad_norm": 0.11013180017471313, "learning_rate": 1.2162179281217211e-05, "loss": 0.0306, "step": 125530 }, { "epoch": 2.5086926981335678, "grad_norm": 0.09007483720779419, "learning_rate": 1.2155424663129488e-05, "loss": 0.0284, "step": 125540 }, { "epoch": 2.5088925302745695, "grad_norm": 0.10497407615184784, "learning_rate": 1.2148671661692518e-05, "loss": 0.0234, "step": 125550 }, { "epoch": 2.5090923624155708, "grad_norm": 0.1136656403541565, "learning_rate": 1.214192027719473e-05, "loss": 0.0221, "step": 125560 }, { "epoch": 2.5092921945565725, "grad_norm": 0.16698245704174042, "learning_rate": 1.2135170509924571e-05, "loss": 0.0265, "step": 125570 }, { "epoch": 2.5094920266975738, "grad_norm": 0.13622571527957916, "learning_rate": 1.2128422360170384e-05, "loss": 0.0283, "step": 125580 }, { "epoch": 2.5096918588385755, "grad_norm": 0.13692840933799744, "learning_rate": 1.2121675828220403e-05, "loss": 0.0248, "step": 125590 }, { "epoch": 2.509891690979577, "grad_norm": 0.08300180733203888, "learning_rate": 1.2114930914362838e-05, "loss": 0.0243, "step": 125600 }, { "epoch": 2.510091523120579, "grad_norm": 0.1082647293806076, "learning_rate": 1.2108187618885853e-05, "loss": 0.0196, "step": 125610 }, { "epoch": 2.51029135526158, "grad_norm": 0.1144464984536171, "learning_rate": 1.2101445942077466e-05, "loss": 0.0297, "step": 125620 }, { "epoch": 2.510491187402582, "grad_norm": 0.14508932828903198, "learning_rate": 1.2094705884225693e-05, "loss": 0.0328, "step": 125630 }, { "epoch": 2.510691019543583, "grad_norm": 0.12188072502613068, "learning_rate": 1.2087967445618443e-05, "loss": 0.0329, "step": 125640 }, { "epoch": 2.510890851684585, "grad_norm": 0.1205352172255516, "learning_rate": 1.2081230626543582e-05, "loss": 0.019, "step": 125650 }, { "epoch": 2.5110906838255866, "grad_norm": 0.10640818625688553, "learning_rate": 1.207449542728889e-05, "loss": 0.0205, "step": 125660 }, { "epoch": 2.511290515966588, "grad_norm": 0.0888596922159195, "learning_rate": 1.206776184814208e-05, "loss": 0.0247, "step": 125670 }, { "epoch": 2.5114903481075896, "grad_norm": 0.15160618722438812, "learning_rate": 1.2061029889390818e-05, "loss": 0.0247, "step": 125680 }, { "epoch": 2.511690180248591, "grad_norm": 0.11533426493406296, "learning_rate": 1.2054299551322629e-05, "loss": 0.0177, "step": 125690 }, { "epoch": 2.5118900123895926, "grad_norm": 0.1376742273569107, "learning_rate": 1.204757083422507e-05, "loss": 0.0248, "step": 125700 }, { "epoch": 2.5120898445305944, "grad_norm": 0.1977008730173111, "learning_rate": 1.2040843738385582e-05, "loss": 0.0225, "step": 125710 }, { "epoch": 2.512289676671596, "grad_norm": 0.08067678660154343, "learning_rate": 1.2034118264091498e-05, "loss": 0.0285, "step": 125720 }, { "epoch": 2.5124895088125974, "grad_norm": 0.12576788663864136, "learning_rate": 1.2027394411630138e-05, "loss": 0.0242, "step": 125730 }, { "epoch": 2.512689340953599, "grad_norm": 0.07959886640310287, "learning_rate": 1.2020672181288734e-05, "loss": 0.0268, "step": 125740 }, { "epoch": 2.5128891730946004, "grad_norm": 0.23625822365283966, "learning_rate": 1.2013951573354437e-05, "loss": 0.0242, "step": 125750 }, { "epoch": 2.513089005235602, "grad_norm": 0.10788512229919434, "learning_rate": 1.2007232588114354e-05, "loss": 0.0262, "step": 125760 }, { "epoch": 2.513288837376604, "grad_norm": 0.10858375579118729, "learning_rate": 1.2000515225855497e-05, "loss": 0.0252, "step": 125770 }, { "epoch": 2.513488669517605, "grad_norm": 0.12887251377105713, "learning_rate": 1.199379948686482e-05, "loss": 0.0295, "step": 125780 }, { "epoch": 2.513688501658607, "grad_norm": 0.11016202718019485, "learning_rate": 1.1987085371429207e-05, "loss": 0.0211, "step": 125790 }, { "epoch": 2.513888333799608, "grad_norm": 0.16060905158519745, "learning_rate": 1.1980372879835484e-05, "loss": 0.0246, "step": 125800 }, { "epoch": 2.51408816594061, "grad_norm": 0.15755340456962585, "learning_rate": 1.1973662012370396e-05, "loss": 0.0237, "step": 125810 }, { "epoch": 2.5142879980816115, "grad_norm": 0.1099664717912674, "learning_rate": 1.1966952769320594e-05, "loss": 0.0186, "step": 125820 }, { "epoch": 2.5144878302226132, "grad_norm": 0.13097821176052094, "learning_rate": 1.1960245150972687e-05, "loss": 0.0199, "step": 125830 }, { "epoch": 2.5146876623636145, "grad_norm": 0.11222927272319794, "learning_rate": 1.1953539157613241e-05, "loss": 0.0235, "step": 125840 }, { "epoch": 2.5148874945046162, "grad_norm": 0.14529968798160553, "learning_rate": 1.1946834789528728e-05, "loss": 0.0239, "step": 125850 }, { "epoch": 2.5150873266456175, "grad_norm": 0.12338650226593018, "learning_rate": 1.194013204700551e-05, "loss": 0.0259, "step": 125860 }, { "epoch": 2.5152871587866192, "grad_norm": 0.06825518608093262, "learning_rate": 1.1933430930329914e-05, "loss": 0.0283, "step": 125870 }, { "epoch": 2.515486990927621, "grad_norm": 0.1236434057354927, "learning_rate": 1.1926731439788246e-05, "loss": 0.0231, "step": 125880 }, { "epoch": 2.5156868230686222, "grad_norm": 0.1792682558298111, "learning_rate": 1.1920033575666662e-05, "loss": 0.0256, "step": 125890 }, { "epoch": 2.515886655209624, "grad_norm": 0.0899660512804985, "learning_rate": 1.1913337338251286e-05, "loss": 0.0195, "step": 125900 }, { "epoch": 2.5160864873506252, "grad_norm": 0.12725579738616943, "learning_rate": 1.1906642727828166e-05, "loss": 0.032, "step": 125910 }, { "epoch": 2.516286319491627, "grad_norm": 0.1328294724225998, "learning_rate": 1.1899949744683292e-05, "loss": 0.0272, "step": 125920 }, { "epoch": 2.5164861516326287, "grad_norm": 0.10025712102651596, "learning_rate": 1.1893258389102574e-05, "loss": 0.0283, "step": 125930 }, { "epoch": 2.5166859837736304, "grad_norm": 0.12210892885923386, "learning_rate": 1.188656866137185e-05, "loss": 0.0204, "step": 125940 }, { "epoch": 2.5168858159146317, "grad_norm": 0.10682240128517151, "learning_rate": 1.1879880561776913e-05, "loss": 0.0225, "step": 125950 }, { "epoch": 2.5170856480556334, "grad_norm": 0.11403627693653107, "learning_rate": 1.1873194090603423e-05, "loss": 0.0216, "step": 125960 }, { "epoch": 2.5172854801966347, "grad_norm": 0.1385660469532013, "learning_rate": 1.1866509248137054e-05, "loss": 0.0274, "step": 125970 }, { "epoch": 2.5174853123376364, "grad_norm": 0.13058751821517944, "learning_rate": 1.1859826034663369e-05, "loss": 0.0177, "step": 125980 }, { "epoch": 2.517685144478638, "grad_norm": 0.11147676408290863, "learning_rate": 1.1853144450467834e-05, "loss": 0.0251, "step": 125990 }, { "epoch": 2.5178849766196394, "grad_norm": 0.15341073274612427, "learning_rate": 1.184646449583589e-05, "loss": 0.0201, "step": 126000 }, { "epoch": 2.518084808760641, "grad_norm": 0.09399525821208954, "learning_rate": 1.1839786171052896e-05, "loss": 0.0217, "step": 126010 }, { "epoch": 2.5182846409016424, "grad_norm": 0.09365251660346985, "learning_rate": 1.183310947640413e-05, "loss": 0.0296, "step": 126020 }, { "epoch": 2.518484473042644, "grad_norm": 0.11819614470005035, "learning_rate": 1.1826434412174813e-05, "loss": 0.0271, "step": 126030 }, { "epoch": 2.518684305183646, "grad_norm": 0.12387112528085709, "learning_rate": 1.1819760978650096e-05, "loss": 0.0236, "step": 126040 }, { "epoch": 2.5188841373246476, "grad_norm": 0.09109117090702057, "learning_rate": 1.181308917611505e-05, "loss": 0.0231, "step": 126050 }, { "epoch": 2.519083969465649, "grad_norm": 0.11162351071834564, "learning_rate": 1.180641900485468e-05, "loss": 0.022, "step": 126060 }, { "epoch": 2.5192838016066506, "grad_norm": 0.13057906925678253, "learning_rate": 1.1799750465153925e-05, "loss": 0.0228, "step": 126070 }, { "epoch": 2.519483633747652, "grad_norm": 0.1368444412946701, "learning_rate": 1.1793083557297668e-05, "loss": 0.0312, "step": 126080 }, { "epoch": 2.5196834658886536, "grad_norm": 0.08685440570116043, "learning_rate": 1.1786418281570683e-05, "loss": 0.0235, "step": 126090 }, { "epoch": 2.5198832980296553, "grad_norm": 0.10325733572244644, "learning_rate": 1.1779754638257689e-05, "loss": 0.0273, "step": 126100 }, { "epoch": 2.5200831301706565, "grad_norm": 0.14833031594753265, "learning_rate": 1.1773092627643395e-05, "loss": 0.0267, "step": 126110 }, { "epoch": 2.5202829623116583, "grad_norm": 0.11363200098276138, "learning_rate": 1.176643225001235e-05, "loss": 0.0185, "step": 126120 }, { "epoch": 2.5204827944526595, "grad_norm": 0.11592815071344376, "learning_rate": 1.1759773505649074e-05, "loss": 0.021, "step": 126130 }, { "epoch": 2.5206826265936613, "grad_norm": 0.0934760794043541, "learning_rate": 1.1753116394838027e-05, "loss": 0.0225, "step": 126140 }, { "epoch": 2.520882458734663, "grad_norm": 0.12729701399803162, "learning_rate": 1.1746460917863583e-05, "loss": 0.0201, "step": 126150 }, { "epoch": 2.5210822908756647, "grad_norm": 0.12188581377267838, "learning_rate": 1.1739807075010061e-05, "loss": 0.0241, "step": 126160 }, { "epoch": 2.521282123016666, "grad_norm": 0.11269310861825943, "learning_rate": 1.173315486656169e-05, "loss": 0.0238, "step": 126170 }, { "epoch": 2.5214819551576677, "grad_norm": 0.08524640649557114, "learning_rate": 1.1726504292802659e-05, "loss": 0.0252, "step": 126180 }, { "epoch": 2.521681787298669, "grad_norm": 0.14621324837207794, "learning_rate": 1.1719855354017029e-05, "loss": 0.0298, "step": 126190 }, { "epoch": 2.5218816194396707, "grad_norm": 0.11392012238502502, "learning_rate": 1.1713208050488866e-05, "loss": 0.0232, "step": 126200 }, { "epoch": 2.5220814515806724, "grad_norm": 0.12497174739837646, "learning_rate": 1.1706562382502134e-05, "loss": 0.0207, "step": 126210 }, { "epoch": 2.5222812837216737, "grad_norm": 0.1328573226928711, "learning_rate": 1.16999183503407e-05, "loss": 0.0225, "step": 126220 }, { "epoch": 2.5224811158626754, "grad_norm": 0.12933851778507233, "learning_rate": 1.1693275954288374e-05, "loss": 0.0269, "step": 126230 }, { "epoch": 2.5226809480036767, "grad_norm": 0.10425490140914917, "learning_rate": 1.1686635194628947e-05, "loss": 0.025, "step": 126240 }, { "epoch": 2.5228807801446784, "grad_norm": 0.1638534963130951, "learning_rate": 1.1679996071646093e-05, "loss": 0.0202, "step": 126250 }, { "epoch": 2.52308061228568, "grad_norm": 0.12816467881202698, "learning_rate": 1.1673358585623395e-05, "loss": 0.031, "step": 126260 }, { "epoch": 2.5232804444266814, "grad_norm": 0.10955853760242462, "learning_rate": 1.166672273684441e-05, "loss": 0.0249, "step": 126270 }, { "epoch": 2.523480276567683, "grad_norm": 0.15074709057807922, "learning_rate": 1.166008852559261e-05, "loss": 0.028, "step": 126280 }, { "epoch": 2.523680108708685, "grad_norm": 0.13300447165966034, "learning_rate": 1.16534559521514e-05, "loss": 0.0207, "step": 126290 }, { "epoch": 2.523879940849686, "grad_norm": 0.11137622594833374, "learning_rate": 1.1646825016804102e-05, "loss": 0.0315, "step": 126300 }, { "epoch": 2.524079772990688, "grad_norm": 0.13843241333961487, "learning_rate": 1.164019571983398e-05, "loss": 0.0196, "step": 126310 }, { "epoch": 2.5242796051316896, "grad_norm": 0.12004269659519196, "learning_rate": 1.163356806152423e-05, "loss": 0.0194, "step": 126320 }, { "epoch": 2.524479437272691, "grad_norm": 0.11829639226198196, "learning_rate": 1.1626942042157974e-05, "loss": 0.0229, "step": 126330 }, { "epoch": 2.5246792694136926, "grad_norm": 0.13367463648319244, "learning_rate": 1.1620317662018259e-05, "loss": 0.022, "step": 126340 }, { "epoch": 2.524879101554694, "grad_norm": 0.10779352486133575, "learning_rate": 1.161369492138808e-05, "loss": 0.026, "step": 126350 }, { "epoch": 2.5250789336956956, "grad_norm": 0.12298613041639328, "learning_rate": 1.1607073820550324e-05, "loss": 0.0215, "step": 126360 }, { "epoch": 2.5252787658366973, "grad_norm": 0.09754060208797455, "learning_rate": 1.1600454359787826e-05, "loss": 0.0241, "step": 126370 }, { "epoch": 2.5254785979776986, "grad_norm": 0.11638442426919937, "learning_rate": 1.1593836539383407e-05, "loss": 0.0284, "step": 126380 }, { "epoch": 2.5256784301187003, "grad_norm": 0.09264563024044037, "learning_rate": 1.1587220359619728e-05, "loss": 0.0207, "step": 126390 }, { "epoch": 2.5258782622597016, "grad_norm": 0.2359379082918167, "learning_rate": 1.1580605820779423e-05, "loss": 0.0161, "step": 126400 }, { "epoch": 2.5260780944007033, "grad_norm": 0.10795042663812637, "learning_rate": 1.1573992923145056e-05, "loss": 0.0223, "step": 126410 }, { "epoch": 2.526277926541705, "grad_norm": 0.25978994369506836, "learning_rate": 1.1567381666999122e-05, "loss": 0.0219, "step": 126420 }, { "epoch": 2.5264777586827067, "grad_norm": 0.1300286203622818, "learning_rate": 1.156077205262404e-05, "loss": 0.026, "step": 126430 }, { "epoch": 2.526677590823708, "grad_norm": 0.10440121591091156, "learning_rate": 1.1554164080302166e-05, "loss": 0.0286, "step": 126440 }, { "epoch": 2.5268774229647097, "grad_norm": 0.1070861965417862, "learning_rate": 1.1547557750315785e-05, "loss": 0.0243, "step": 126450 }, { "epoch": 2.527077255105711, "grad_norm": 0.16584216058254242, "learning_rate": 1.1540953062947069e-05, "loss": 0.0258, "step": 126460 }, { "epoch": 2.5272770872467127, "grad_norm": 0.13235484063625336, "learning_rate": 1.1534350018478202e-05, "loss": 0.0254, "step": 126470 }, { "epoch": 2.5274769193877145, "grad_norm": 0.1285805106163025, "learning_rate": 1.1527748617191253e-05, "loss": 0.029, "step": 126480 }, { "epoch": 2.5276767515287157, "grad_norm": 0.14051488041877747, "learning_rate": 1.1521148859368192e-05, "loss": 0.0252, "step": 126490 }, { "epoch": 2.5278765836697175, "grad_norm": 0.1150066927075386, "learning_rate": 1.1514550745290953e-05, "loss": 0.02, "step": 126500 }, { "epoch": 2.5280764158107187, "grad_norm": 0.1275688260793686, "learning_rate": 1.1507954275241428e-05, "loss": 0.0273, "step": 126510 }, { "epoch": 2.5282762479517205, "grad_norm": 0.11139236390590668, "learning_rate": 1.1501359449501375e-05, "loss": 0.0296, "step": 126520 }, { "epoch": 2.528476080092722, "grad_norm": 0.18342779576778412, "learning_rate": 1.1494766268352519e-05, "loss": 0.0229, "step": 126530 }, { "epoch": 2.528675912233724, "grad_norm": 0.13660776615142822, "learning_rate": 1.1488174732076517e-05, "loss": 0.0262, "step": 126540 }, { "epoch": 2.528875744374725, "grad_norm": 0.11077435314655304, "learning_rate": 1.1481584840954933e-05, "loss": 0.0163, "step": 126550 }, { "epoch": 2.529075576515727, "grad_norm": 0.15676575899124146, "learning_rate": 1.1474996595269294e-05, "loss": 0.0288, "step": 126560 }, { "epoch": 2.529275408656728, "grad_norm": 0.18684332072734833, "learning_rate": 1.1468409995301021e-05, "loss": 0.0301, "step": 126570 }, { "epoch": 2.52947524079773, "grad_norm": 0.148373082280159, "learning_rate": 1.1461825041331504e-05, "loss": 0.0271, "step": 126580 }, { "epoch": 2.5296750729387316, "grad_norm": 0.12113133817911148, "learning_rate": 1.1455241733641997e-05, "loss": 0.0199, "step": 126590 }, { "epoch": 2.529874905079733, "grad_norm": 0.10491590201854706, "learning_rate": 1.1448660072513767e-05, "loss": 0.0338, "step": 126600 }, { "epoch": 2.5300747372207346, "grad_norm": 0.11050280928611755, "learning_rate": 1.1442080058227973e-05, "loss": 0.0206, "step": 126610 }, { "epoch": 2.530274569361736, "grad_norm": 0.11738403886556625, "learning_rate": 1.1435501691065669e-05, "loss": 0.0234, "step": 126620 }, { "epoch": 2.5304744015027376, "grad_norm": 0.15491385757923126, "learning_rate": 1.1428924971307892e-05, "loss": 0.0244, "step": 126630 }, { "epoch": 2.5306742336437393, "grad_norm": 0.13621778786182404, "learning_rate": 1.1422349899235569e-05, "loss": 0.0181, "step": 126640 }, { "epoch": 2.530874065784741, "grad_norm": 0.08152474462985992, "learning_rate": 1.1415776475129614e-05, "loss": 0.0193, "step": 126650 }, { "epoch": 2.5310738979257423, "grad_norm": 0.10401042550802231, "learning_rate": 1.1409204699270792e-05, "loss": 0.0251, "step": 126660 }, { "epoch": 2.531273730066744, "grad_norm": 0.10037790983915329, "learning_rate": 1.1402634571939853e-05, "loss": 0.0265, "step": 126670 }, { "epoch": 2.5314735622077453, "grad_norm": 0.1141313835978508, "learning_rate": 1.1396066093417457e-05, "loss": 0.0253, "step": 126680 }, { "epoch": 2.531673394348747, "grad_norm": 0.10196304321289062, "learning_rate": 1.1389499263984205e-05, "loss": 0.0235, "step": 126690 }, { "epoch": 2.5318732264897488, "grad_norm": 0.10824903100728989, "learning_rate": 1.138293408392061e-05, "loss": 0.0243, "step": 126700 }, { "epoch": 2.53207305863075, "grad_norm": 0.15311263501644135, "learning_rate": 1.1376370553507132e-05, "loss": 0.022, "step": 126710 }, { "epoch": 2.5322728907717518, "grad_norm": 0.11425560712814331, "learning_rate": 1.1369808673024146e-05, "loss": 0.0293, "step": 126720 }, { "epoch": 2.532472722912753, "grad_norm": 0.1692519336938858, "learning_rate": 1.1363248442751973e-05, "loss": 0.0242, "step": 126730 }, { "epoch": 2.5326725550537548, "grad_norm": 0.17322063446044922, "learning_rate": 1.1356689862970844e-05, "loss": 0.0241, "step": 126740 }, { "epoch": 2.5328723871947565, "grad_norm": 0.12679365277290344, "learning_rate": 1.135013293396095e-05, "loss": 0.0229, "step": 126750 }, { "epoch": 2.533072219335758, "grad_norm": 0.09118221700191498, "learning_rate": 1.1343577656002364e-05, "loss": 0.0268, "step": 126760 }, { "epoch": 2.5332720514767595, "grad_norm": 0.14346174895763397, "learning_rate": 1.133702402937511e-05, "loss": 0.0244, "step": 126770 }, { "epoch": 2.533471883617761, "grad_norm": 0.08429322391748428, "learning_rate": 1.1330472054359188e-05, "loss": 0.02, "step": 126780 }, { "epoch": 2.5336717157587625, "grad_norm": 0.1373995989561081, "learning_rate": 1.1323921731234449e-05, "loss": 0.0223, "step": 126790 }, { "epoch": 2.533871547899764, "grad_norm": 0.09919492155313492, "learning_rate": 1.1317373060280728e-05, "loss": 0.0314, "step": 126800 }, { "epoch": 2.534071380040766, "grad_norm": 0.14095844328403473, "learning_rate": 1.1310826041777762e-05, "loss": 0.0247, "step": 126810 }, { "epoch": 2.534271212181767, "grad_norm": 0.10517965257167816, "learning_rate": 1.1304280676005235e-05, "loss": 0.0241, "step": 126820 }, { "epoch": 2.534471044322769, "grad_norm": 0.11793794482946396, "learning_rate": 1.1297736963242756e-05, "loss": 0.0259, "step": 126830 }, { "epoch": 2.53467087646377, "grad_norm": 0.13266797363758087, "learning_rate": 1.1291194903769846e-05, "loss": 0.0217, "step": 126840 }, { "epoch": 2.534870708604772, "grad_norm": 0.11366438120603561, "learning_rate": 1.1284654497865998e-05, "loss": 0.0209, "step": 126850 }, { "epoch": 2.5350705407457736, "grad_norm": 0.13637970387935638, "learning_rate": 1.1278115745810564e-05, "loss": 0.0234, "step": 126860 }, { "epoch": 2.5352703728867754, "grad_norm": 0.13770583271980286, "learning_rate": 1.1271578647882901e-05, "loss": 0.0156, "step": 126870 }, { "epoch": 2.5354702050277766, "grad_norm": 0.17499659955501556, "learning_rate": 1.1265043204362263e-05, "loss": 0.0279, "step": 126880 }, { "epoch": 2.5356700371687784, "grad_norm": 0.13813498616218567, "learning_rate": 1.1258509415527813e-05, "loss": 0.0207, "step": 126890 }, { "epoch": 2.5358698693097796, "grad_norm": 0.11162015050649643, "learning_rate": 1.1251977281658654e-05, "loss": 0.0288, "step": 126900 }, { "epoch": 2.5360697014507814, "grad_norm": 0.10209858417510986, "learning_rate": 1.1245446803033872e-05, "loss": 0.0221, "step": 126910 }, { "epoch": 2.536269533591783, "grad_norm": 0.09964307397603989, "learning_rate": 1.123891797993239e-05, "loss": 0.0212, "step": 126920 }, { "epoch": 2.5364693657327844, "grad_norm": 0.11302298307418823, "learning_rate": 1.1232390812633125e-05, "loss": 0.0229, "step": 126930 }, { "epoch": 2.536669197873786, "grad_norm": 0.1459995061159134, "learning_rate": 1.1225865301414912e-05, "loss": 0.0315, "step": 126940 }, { "epoch": 2.5368690300147874, "grad_norm": 0.11910805851221085, "learning_rate": 1.1219341446556497e-05, "loss": 0.0266, "step": 126950 }, { "epoch": 2.537068862155789, "grad_norm": 0.11128754168748856, "learning_rate": 1.1212819248336581e-05, "loss": 0.0211, "step": 126960 }, { "epoch": 2.537268694296791, "grad_norm": 0.09031616896390915, "learning_rate": 1.1206298707033764e-05, "loss": 0.0281, "step": 126970 }, { "epoch": 2.5374685264377925, "grad_norm": 0.11264439672231674, "learning_rate": 1.119977982292662e-05, "loss": 0.0233, "step": 126980 }, { "epoch": 2.537668358578794, "grad_norm": 0.1641201376914978, "learning_rate": 1.1193262596293574e-05, "loss": 0.0253, "step": 126990 }, { "epoch": 2.5378681907197955, "grad_norm": 0.09759242832660675, "learning_rate": 1.1186747027413069e-05, "loss": 0.0303, "step": 127000 }, { "epoch": 2.538068022860797, "grad_norm": 0.08685117959976196, "learning_rate": 1.1180233116563448e-05, "loss": 0.0174, "step": 127010 }, { "epoch": 2.5382678550017985, "grad_norm": 0.11972053349018097, "learning_rate": 1.117372086402294e-05, "loss": 0.0186, "step": 127020 }, { "epoch": 2.5384676871428002, "grad_norm": 0.09586463868618011, "learning_rate": 1.116721027006975e-05, "loss": 0.0258, "step": 127030 }, { "epoch": 2.5386675192838015, "grad_norm": 0.08515503257513046, "learning_rate": 1.1160701334981982e-05, "loss": 0.0265, "step": 127040 }, { "epoch": 2.5388673514248032, "grad_norm": 0.13080209493637085, "learning_rate": 1.1154194059037732e-05, "loss": 0.0283, "step": 127050 }, { "epoch": 2.5390671835658045, "grad_norm": 0.09167835861444473, "learning_rate": 1.1147688442514937e-05, "loss": 0.0219, "step": 127060 }, { "epoch": 2.5392670157068062, "grad_norm": 0.1266922801733017, "learning_rate": 1.114118448569152e-05, "loss": 0.0258, "step": 127070 }, { "epoch": 2.539466847847808, "grad_norm": 0.1297609955072403, "learning_rate": 1.1134682188845314e-05, "loss": 0.0245, "step": 127080 }, { "epoch": 2.5396666799888092, "grad_norm": 0.09504621475934982, "learning_rate": 1.1128181552254085e-05, "loss": 0.0272, "step": 127090 }, { "epoch": 2.539866512129811, "grad_norm": 0.11861182749271393, "learning_rate": 1.1121682576195536e-05, "loss": 0.0243, "step": 127100 }, { "epoch": 2.5400663442708127, "grad_norm": 0.12946462631225586, "learning_rate": 1.111518526094728e-05, "loss": 0.0296, "step": 127110 }, { "epoch": 2.540266176411814, "grad_norm": 0.09086833894252777, "learning_rate": 1.1108689606786888e-05, "loss": 0.0226, "step": 127120 }, { "epoch": 2.5404660085528157, "grad_norm": 0.1202034056186676, "learning_rate": 1.1102195613991806e-05, "loss": 0.039, "step": 127130 }, { "epoch": 2.5406658406938174, "grad_norm": 0.13450977206230164, "learning_rate": 1.1095703282839486e-05, "loss": 0.0236, "step": 127140 }, { "epoch": 2.5408656728348187, "grad_norm": 0.11097732931375504, "learning_rate": 1.1089212613607264e-05, "loss": 0.0241, "step": 127150 }, { "epoch": 2.5410655049758204, "grad_norm": 0.12090931832790375, "learning_rate": 1.1082723606572382e-05, "loss": 0.0311, "step": 127160 }, { "epoch": 2.5412653371168217, "grad_norm": 0.12571701407432556, "learning_rate": 1.107623626201204e-05, "loss": 0.0312, "step": 127170 }, { "epoch": 2.5414651692578234, "grad_norm": 0.13392457365989685, "learning_rate": 1.1069750580203413e-05, "loss": 0.0207, "step": 127180 }, { "epoch": 2.541665001398825, "grad_norm": 0.12465143948793411, "learning_rate": 1.1063266561423507e-05, "loss": 0.0226, "step": 127190 }, { "epoch": 2.5418648335398264, "grad_norm": 0.1008494570851326, "learning_rate": 1.1056784205949322e-05, "loss": 0.0194, "step": 127200 }, { "epoch": 2.542064665680828, "grad_norm": 0.11783867329359055, "learning_rate": 1.105030351405778e-05, "loss": 0.019, "step": 127210 }, { "epoch": 2.5422644978218294, "grad_norm": 0.17776493728160858, "learning_rate": 1.1043824486025711e-05, "loss": 0.028, "step": 127220 }, { "epoch": 2.542464329962831, "grad_norm": 0.15968213975429535, "learning_rate": 1.1037347122129903e-05, "loss": 0.028, "step": 127230 }, { "epoch": 2.542664162103833, "grad_norm": 0.1272760033607483, "learning_rate": 1.1030871422647044e-05, "loss": 0.0252, "step": 127240 }, { "epoch": 2.5428639942448346, "grad_norm": 0.11668166518211365, "learning_rate": 1.1024397387853785e-05, "loss": 0.0259, "step": 127250 }, { "epoch": 2.543063826385836, "grad_norm": 0.14933548867702484, "learning_rate": 1.1017925018026642e-05, "loss": 0.0297, "step": 127260 }, { "epoch": 2.5432636585268376, "grad_norm": 0.16627644002437592, "learning_rate": 1.1011454313442144e-05, "loss": 0.023, "step": 127270 }, { "epoch": 2.543463490667839, "grad_norm": 0.14210890233516693, "learning_rate": 1.1004985274376705e-05, "loss": 0.0277, "step": 127280 }, { "epoch": 2.5436633228088406, "grad_norm": 0.1488925963640213, "learning_rate": 1.0998517901106653e-05, "loss": 0.0361, "step": 127290 }, { "epoch": 2.5438631549498423, "grad_norm": 0.13984741270542145, "learning_rate": 1.0992052193908259e-05, "loss": 0.0234, "step": 127300 }, { "epoch": 2.5440629870908436, "grad_norm": 0.1693749576807022, "learning_rate": 1.0985588153057741e-05, "loss": 0.0194, "step": 127310 }, { "epoch": 2.5442628192318453, "grad_norm": 0.11075577884912491, "learning_rate": 1.0979125778831223e-05, "loss": 0.0233, "step": 127320 }, { "epoch": 2.5444626513728466, "grad_norm": 0.10839925706386566, "learning_rate": 1.0972665071504768e-05, "loss": 0.0244, "step": 127330 }, { "epoch": 2.5446624835138483, "grad_norm": 0.12859410047531128, "learning_rate": 1.0966206031354364e-05, "loss": 0.0217, "step": 127340 }, { "epoch": 2.54486231565485, "grad_norm": 0.11372894048690796, "learning_rate": 1.0959748658655933e-05, "loss": 0.0238, "step": 127350 }, { "epoch": 2.5450621477958517, "grad_norm": 0.10758864134550095, "learning_rate": 1.0953292953685317e-05, "loss": 0.0293, "step": 127360 }, { "epoch": 2.545261979936853, "grad_norm": 0.17584720253944397, "learning_rate": 1.09468389167183e-05, "loss": 0.0282, "step": 127370 }, { "epoch": 2.5454618120778547, "grad_norm": 0.11202060431241989, "learning_rate": 1.094038654803059e-05, "loss": 0.0242, "step": 127380 }, { "epoch": 2.545661644218856, "grad_norm": 0.14457111060619354, "learning_rate": 1.0933935847897803e-05, "loss": 0.0365, "step": 127390 }, { "epoch": 2.5458614763598577, "grad_norm": 0.1374584585428238, "learning_rate": 1.0927486816595494e-05, "loss": 0.0294, "step": 127400 }, { "epoch": 2.5460613085008594, "grad_norm": 0.2247827798128128, "learning_rate": 1.0921039454399196e-05, "loss": 0.0304, "step": 127410 }, { "epoch": 2.5462611406418607, "grad_norm": 0.10931110382080078, "learning_rate": 1.091459376158429e-05, "loss": 0.0225, "step": 127420 }, { "epoch": 2.5464609727828624, "grad_norm": 0.10110551118850708, "learning_rate": 1.0908149738426138e-05, "loss": 0.0283, "step": 127430 }, { "epoch": 2.5466608049238637, "grad_norm": 0.1277874857187271, "learning_rate": 1.09017073852e-05, "loss": 0.0249, "step": 127440 }, { "epoch": 2.5468606370648654, "grad_norm": 0.1340598165988922, "learning_rate": 1.0895266702181122e-05, "loss": 0.0204, "step": 127450 }, { "epoch": 2.547060469205867, "grad_norm": 0.1360914409160614, "learning_rate": 1.0888827689644599e-05, "loss": 0.0223, "step": 127460 }, { "epoch": 2.547260301346869, "grad_norm": 0.11861209571361542, "learning_rate": 1.0882390347865512e-05, "loss": 0.0276, "step": 127470 }, { "epoch": 2.54746013348787, "grad_norm": 0.10870479047298431, "learning_rate": 1.0875954677118844e-05, "loss": 0.0233, "step": 127480 }, { "epoch": 2.547659965628872, "grad_norm": 0.11042755097150803, "learning_rate": 1.0869520677679524e-05, "loss": 0.0216, "step": 127490 }, { "epoch": 2.547859797769873, "grad_norm": 0.14553016424179077, "learning_rate": 1.0863088349822386e-05, "loss": 0.0291, "step": 127500 }, { "epoch": 2.548059629910875, "grad_norm": 0.11836099624633789, "learning_rate": 1.0856657693822225e-05, "loss": 0.025, "step": 127510 }, { "epoch": 2.5482594620518766, "grad_norm": 0.14102862775325775, "learning_rate": 1.0850228709953753e-05, "loss": 0.0264, "step": 127520 }, { "epoch": 2.548459294192878, "grad_norm": 0.1365133672952652, "learning_rate": 1.084380139849156e-05, "loss": 0.0233, "step": 127530 }, { "epoch": 2.5486591263338796, "grad_norm": 0.12622253596782684, "learning_rate": 1.083737575971025e-05, "loss": 0.0308, "step": 127540 }, { "epoch": 2.548858958474881, "grad_norm": 0.10932252556085587, "learning_rate": 1.083095179388432e-05, "loss": 0.0297, "step": 127550 }, { "epoch": 2.5490587906158826, "grad_norm": 0.10205941647291183, "learning_rate": 1.0824529501288161e-05, "loss": 0.0265, "step": 127560 }, { "epoch": 2.5492586227568843, "grad_norm": 0.11626841127872467, "learning_rate": 1.0818108882196132e-05, "loss": 0.03, "step": 127570 }, { "epoch": 2.549458454897886, "grad_norm": 0.11321134865283966, "learning_rate": 1.0811689936882513e-05, "loss": 0.0185, "step": 127580 }, { "epoch": 2.5496582870388873, "grad_norm": 0.14034341275691986, "learning_rate": 1.0805272665621507e-05, "loss": 0.0318, "step": 127590 }, { "epoch": 2.549858119179889, "grad_norm": 0.12098889797925949, "learning_rate": 1.0798857068687252e-05, "loss": 0.0249, "step": 127600 }, { "epoch": 2.5500579513208903, "grad_norm": 0.14953574538230896, "learning_rate": 1.079244314635381e-05, "loss": 0.0292, "step": 127610 }, { "epoch": 2.550257783461892, "grad_norm": 0.10612081736326218, "learning_rate": 1.0786030898895167e-05, "loss": 0.0242, "step": 127620 }, { "epoch": 2.5504576156028937, "grad_norm": 0.13905562460422516, "learning_rate": 1.0779620326585244e-05, "loss": 0.0258, "step": 127630 }, { "epoch": 2.550657447743895, "grad_norm": 0.0892886370420456, "learning_rate": 1.0773211429697894e-05, "loss": 0.0202, "step": 127640 }, { "epoch": 2.5508572798848967, "grad_norm": 0.08869879692792892, "learning_rate": 1.0766804208506903e-05, "loss": 0.0221, "step": 127650 }, { "epoch": 2.551057112025898, "grad_norm": 0.14000160992145538, "learning_rate": 1.0760398663285948e-05, "loss": 0.0207, "step": 127660 }, { "epoch": 2.5512569441668997, "grad_norm": 0.10650789737701416, "learning_rate": 1.0753994794308664e-05, "loss": 0.0171, "step": 127670 }, { "epoch": 2.5514567763079015, "grad_norm": 0.11694994568824768, "learning_rate": 1.0747592601848649e-05, "loss": 0.0196, "step": 127680 }, { "epoch": 2.551656608448903, "grad_norm": 0.14234842360019684, "learning_rate": 1.0741192086179358e-05, "loss": 0.0266, "step": 127690 }, { "epoch": 2.5518564405899045, "grad_norm": 0.1348087042570114, "learning_rate": 1.073479324757422e-05, "loss": 0.0205, "step": 127700 }, { "epoch": 2.552056272730906, "grad_norm": 0.11862681061029434, "learning_rate": 1.0728396086306585e-05, "loss": 0.0207, "step": 127710 }, { "epoch": 2.5522561048719075, "grad_norm": 0.10455717146396637, "learning_rate": 1.0722000602649717e-05, "loss": 0.0216, "step": 127720 }, { "epoch": 2.552455937012909, "grad_norm": 0.1376534253358841, "learning_rate": 1.0715606796876838e-05, "loss": 0.0312, "step": 127730 }, { "epoch": 2.552655769153911, "grad_norm": 0.12383067607879639, "learning_rate": 1.0709214669261065e-05, "loss": 0.0209, "step": 127740 }, { "epoch": 2.552855601294912, "grad_norm": 0.13570764660835266, "learning_rate": 1.070282422007547e-05, "loss": 0.0298, "step": 127750 }, { "epoch": 2.553055433435914, "grad_norm": 0.15388962626457214, "learning_rate": 1.0696435449593012e-05, "loss": 0.0233, "step": 127760 }, { "epoch": 2.553255265576915, "grad_norm": 0.11318907886743546, "learning_rate": 1.069004835808664e-05, "loss": 0.0217, "step": 127770 }, { "epoch": 2.553455097717917, "grad_norm": 0.10143651068210602, "learning_rate": 1.0683662945829198e-05, "loss": 0.038, "step": 127780 }, { "epoch": 2.5536549298589186, "grad_norm": 0.1408708393573761, "learning_rate": 1.0677279213093438e-05, "loss": 0.042, "step": 127790 }, { "epoch": 2.5538547619999203, "grad_norm": 0.12687069177627563, "learning_rate": 1.067089716015206e-05, "loss": 0.0194, "step": 127800 }, { "epoch": 2.5540545941409216, "grad_norm": 0.11032616347074509, "learning_rate": 1.0664516787277729e-05, "loss": 0.0269, "step": 127810 }, { "epoch": 2.5542544262819233, "grad_norm": 0.0988883376121521, "learning_rate": 1.065813809474297e-05, "loss": 0.0193, "step": 127820 }, { "epoch": 2.5544542584229246, "grad_norm": 0.16240182518959045, "learning_rate": 1.0651761082820278e-05, "loss": 0.0214, "step": 127830 }, { "epoch": 2.5546540905639263, "grad_norm": 0.2174983024597168, "learning_rate": 1.064538575178205e-05, "loss": 0.0218, "step": 127840 }, { "epoch": 2.554853922704928, "grad_norm": 0.13369612395763397, "learning_rate": 1.0639012101900675e-05, "loss": 0.0216, "step": 127850 }, { "epoch": 2.5550537548459293, "grad_norm": 0.10305888950824738, "learning_rate": 1.0632640133448379e-05, "loss": 0.0316, "step": 127860 }, { "epoch": 2.555253586986931, "grad_norm": 0.10198093950748444, "learning_rate": 1.0626269846697379e-05, "loss": 0.0323, "step": 127870 }, { "epoch": 2.5554534191279323, "grad_norm": 0.10408805310726166, "learning_rate": 1.06199012419198e-05, "loss": 0.0276, "step": 127880 }, { "epoch": 2.555653251268934, "grad_norm": 0.09351956099271774, "learning_rate": 1.0613534319387692e-05, "loss": 0.0293, "step": 127890 }, { "epoch": 2.5558530834099358, "grad_norm": 0.15416774153709412, "learning_rate": 1.0607169079373048e-05, "loss": 0.0246, "step": 127900 }, { "epoch": 2.556052915550937, "grad_norm": 0.126479834318161, "learning_rate": 1.0600805522147772e-05, "loss": 0.0282, "step": 127910 }, { "epoch": 2.5562527476919388, "grad_norm": 0.11097197234630585, "learning_rate": 1.0594443647983715e-05, "loss": 0.0225, "step": 127920 }, { "epoch": 2.55645257983294, "grad_norm": 0.15315981209278107, "learning_rate": 1.0588083457152614e-05, "loss": 0.0215, "step": 127930 }, { "epoch": 2.5566524119739418, "grad_norm": 0.19976548850536346, "learning_rate": 1.0581724949926198e-05, "loss": 0.0236, "step": 127940 }, { "epoch": 2.5568522441149435, "grad_norm": 0.09092005342245102, "learning_rate": 1.057536812657609e-05, "loss": 0.0193, "step": 127950 }, { "epoch": 2.557052076255945, "grad_norm": 0.11231043934822083, "learning_rate": 1.0569012987373816e-05, "loss": 0.0225, "step": 127960 }, { "epoch": 2.5572519083969465, "grad_norm": 0.11625924706459045, "learning_rate": 1.0562659532590874e-05, "loss": 0.0197, "step": 127970 }, { "epoch": 2.557451740537948, "grad_norm": 0.17622293531894684, "learning_rate": 1.0556307762498663e-05, "loss": 0.0246, "step": 127980 }, { "epoch": 2.5576515726789495, "grad_norm": 0.1474425047636032, "learning_rate": 1.0549957677368527e-05, "loss": 0.0238, "step": 127990 }, { "epoch": 2.557851404819951, "grad_norm": 0.10353751480579376, "learning_rate": 1.0543609277471727e-05, "loss": 0.0185, "step": 128000 }, { "epoch": 2.558051236960953, "grad_norm": 0.10914795845746994, "learning_rate": 1.0537262563079452e-05, "loss": 0.0298, "step": 128010 }, { "epoch": 2.558251069101954, "grad_norm": 0.11233624815940857, "learning_rate": 1.053091753446283e-05, "loss": 0.0219, "step": 128020 }, { "epoch": 2.558450901242956, "grad_norm": 0.1437361240386963, "learning_rate": 1.05245741918929e-05, "loss": 0.0278, "step": 128030 }, { "epoch": 2.558650733383957, "grad_norm": 0.11065985262393951, "learning_rate": 1.0518232535640644e-05, "loss": 0.0273, "step": 128040 }, { "epoch": 2.558850565524959, "grad_norm": 0.10339752584695816, "learning_rate": 1.0511892565976978e-05, "loss": 0.0231, "step": 128050 }, { "epoch": 2.5590503976659607, "grad_norm": 0.07795560359954834, "learning_rate": 1.0505554283172703e-05, "loss": 0.0208, "step": 128060 }, { "epoch": 2.5592502298069624, "grad_norm": 0.09694258868694305, "learning_rate": 1.0499217687498585e-05, "loss": 0.0325, "step": 128070 }, { "epoch": 2.5594500619479637, "grad_norm": 0.1359591782093048, "learning_rate": 1.0492882779225349e-05, "loss": 0.0266, "step": 128080 }, { "epoch": 2.5596498940889654, "grad_norm": 0.14147518575191498, "learning_rate": 1.0486549558623571e-05, "loss": 0.0306, "step": 128090 }, { "epoch": 2.5598497262299666, "grad_norm": 0.15275618433952332, "learning_rate": 1.04802180259638e-05, "loss": 0.033, "step": 128100 }, { "epoch": 2.5600495583709684, "grad_norm": 0.12333858013153076, "learning_rate": 1.0473888181516523e-05, "loss": 0.0219, "step": 128110 }, { "epoch": 2.56024939051197, "grad_norm": 0.10450100153684616, "learning_rate": 1.0467560025552126e-05, "loss": 0.0273, "step": 128120 }, { "epoch": 2.5604492226529714, "grad_norm": 0.12952347099781036, "learning_rate": 1.0461233558340943e-05, "loss": 0.0254, "step": 128130 }, { "epoch": 2.560649054793973, "grad_norm": 0.14484025537967682, "learning_rate": 1.0454908780153227e-05, "loss": 0.027, "step": 128140 }, { "epoch": 2.5608488869349744, "grad_norm": 0.14987456798553467, "learning_rate": 1.0448585691259178e-05, "loss": 0.027, "step": 128150 }, { "epoch": 2.561048719075976, "grad_norm": 0.10856149345636368, "learning_rate": 1.044226429192886e-05, "loss": 0.0226, "step": 128160 }, { "epoch": 2.561248551216978, "grad_norm": 0.13601651787757874, "learning_rate": 1.0435944582432355e-05, "loss": 0.0263, "step": 128170 }, { "epoch": 2.5614483833579795, "grad_norm": 0.10267742723226547, "learning_rate": 1.042962656303963e-05, "loss": 0.0224, "step": 128180 }, { "epoch": 2.561648215498981, "grad_norm": 0.13781943917274475, "learning_rate": 1.0423310234020556e-05, "loss": 0.0304, "step": 128190 }, { "epoch": 2.5618480476399825, "grad_norm": 0.1269099861383438, "learning_rate": 1.0416995595644946e-05, "loss": 0.0286, "step": 128200 }, { "epoch": 2.562047879780984, "grad_norm": 0.10780909657478333, "learning_rate": 1.0410682648182579e-05, "loss": 0.0212, "step": 128210 }, { "epoch": 2.5622477119219855, "grad_norm": 0.1045665442943573, "learning_rate": 1.040437139190314e-05, "loss": 0.0252, "step": 128220 }, { "epoch": 2.5624475440629872, "grad_norm": 0.11247380077838898, "learning_rate": 1.03980618270762e-05, "loss": 0.0197, "step": 128230 }, { "epoch": 2.5626473762039885, "grad_norm": 0.10314954072237015, "learning_rate": 1.0391753953971307e-05, "loss": 0.0243, "step": 128240 }, { "epoch": 2.5628472083449902, "grad_norm": 0.09418972581624985, "learning_rate": 1.038544777285792e-05, "loss": 0.0201, "step": 128250 }, { "epoch": 2.5630470404859915, "grad_norm": 0.13692441582679749, "learning_rate": 1.0379143284005432e-05, "loss": 0.0346, "step": 128260 }, { "epoch": 2.5632468726269932, "grad_norm": 0.1304553896188736, "learning_rate": 1.0372840487683161e-05, "loss": 0.0177, "step": 128270 }, { "epoch": 2.563446704767995, "grad_norm": 0.14113624393939972, "learning_rate": 1.036653938416034e-05, "loss": 0.0192, "step": 128280 }, { "epoch": 2.5636465369089967, "grad_norm": 0.14780476689338684, "learning_rate": 1.0360239973706153e-05, "loss": 0.0226, "step": 128290 }, { "epoch": 2.563846369049998, "grad_norm": 0.0799289271235466, "learning_rate": 1.0353942256589694e-05, "loss": 0.0191, "step": 128300 }, { "epoch": 2.5640462011909997, "grad_norm": 0.09877022355794907, "learning_rate": 1.0347646233079989e-05, "loss": 0.0224, "step": 128310 }, { "epoch": 2.564246033332001, "grad_norm": 0.14773117005825043, "learning_rate": 1.0341351903446e-05, "loss": 0.0312, "step": 128320 }, { "epoch": 2.5644458654730027, "grad_norm": 0.0869927704334259, "learning_rate": 1.0335059267956598e-05, "loss": 0.0204, "step": 128330 }, { "epoch": 2.5646456976140044, "grad_norm": 0.17765849828720093, "learning_rate": 1.0328768326880578e-05, "loss": 0.0275, "step": 128340 }, { "epoch": 2.5648455297550057, "grad_norm": 0.12386492639780045, "learning_rate": 1.0322479080486725e-05, "loss": 0.0227, "step": 128350 }, { "epoch": 2.5650453618960074, "grad_norm": 0.10277390480041504, "learning_rate": 1.0316191529043662e-05, "loss": 0.0243, "step": 128360 }, { "epoch": 2.5652451940370087, "grad_norm": 0.15617267787456512, "learning_rate": 1.0309905672819998e-05, "loss": 0.0244, "step": 128370 }, { "epoch": 2.5654450261780104, "grad_norm": 0.11404974013566971, "learning_rate": 1.030362151208425e-05, "loss": 0.0322, "step": 128380 }, { "epoch": 2.565644858319012, "grad_norm": 0.1138727068901062, "learning_rate": 1.0297339047104865e-05, "loss": 0.025, "step": 128390 }, { "epoch": 2.565844690460014, "grad_norm": 0.10987342149019241, "learning_rate": 1.0291058278150223e-05, "loss": 0.0316, "step": 128400 }, { "epoch": 2.566044522601015, "grad_norm": 0.11610493808984756, "learning_rate": 1.028477920548862e-05, "loss": 0.0253, "step": 128410 }, { "epoch": 2.566244354742017, "grad_norm": 0.09258054196834564, "learning_rate": 1.0278501829388303e-05, "loss": 0.017, "step": 128420 }, { "epoch": 2.566444186883018, "grad_norm": 0.10154854506254196, "learning_rate": 1.0272226150117392e-05, "loss": 0.0219, "step": 128430 }, { "epoch": 2.56664401902402, "grad_norm": 0.13665279746055603, "learning_rate": 1.026595216794401e-05, "loss": 0.0355, "step": 128440 }, { "epoch": 2.5668438511650216, "grad_norm": 0.08110575377941132, "learning_rate": 1.025967988313617e-05, "loss": 0.03, "step": 128450 }, { "epoch": 2.567043683306023, "grad_norm": 0.14261431992053986, "learning_rate": 1.0253409295961791e-05, "loss": 0.0239, "step": 128460 }, { "epoch": 2.5672435154470246, "grad_norm": 0.11427391320466995, "learning_rate": 1.0247140406688733e-05, "loss": 0.0255, "step": 128470 }, { "epoch": 2.567443347588026, "grad_norm": 0.12267649918794632, "learning_rate": 1.024087321558484e-05, "loss": 0.0243, "step": 128480 }, { "epoch": 2.5676431797290276, "grad_norm": 0.1351534128189087, "learning_rate": 1.0234607722917783e-05, "loss": 0.0276, "step": 128490 }, { "epoch": 2.5678430118700293, "grad_norm": 0.10901567339897156, "learning_rate": 1.0228343928955236e-05, "loss": 0.0312, "step": 128500 }, { "epoch": 2.568042844011031, "grad_norm": 0.10884765535593033, "learning_rate": 1.0222081833964775e-05, "loss": 0.0238, "step": 128510 }, { "epoch": 2.5682426761520323, "grad_norm": 0.09392891079187393, "learning_rate": 1.0215821438213902e-05, "loss": 0.0221, "step": 128520 }, { "epoch": 2.568442508293034, "grad_norm": 0.10557930171489716, "learning_rate": 1.0209562741970053e-05, "loss": 0.0237, "step": 128530 }, { "epoch": 2.5686423404340353, "grad_norm": 0.09572073817253113, "learning_rate": 1.0203305745500585e-05, "loss": 0.0202, "step": 128540 }, { "epoch": 2.568842172575037, "grad_norm": 0.1257794201374054, "learning_rate": 1.0197050449072799e-05, "loss": 0.0333, "step": 128550 }, { "epoch": 2.5690420047160387, "grad_norm": 0.12234606593847275, "learning_rate": 1.0190796852953872e-05, "loss": 0.0244, "step": 128560 }, { "epoch": 2.56924183685704, "grad_norm": 0.09914788603782654, "learning_rate": 1.0184544957410985e-05, "loss": 0.0317, "step": 128570 }, { "epoch": 2.5694416689980417, "grad_norm": 0.1061442494392395, "learning_rate": 1.0178294762711204e-05, "loss": 0.02, "step": 128580 }, { "epoch": 2.569641501139043, "grad_norm": 0.0947001501917839, "learning_rate": 1.0172046269121505e-05, "loss": 0.017, "step": 128590 }, { "epoch": 2.5698413332800447, "grad_norm": 0.09816849231719971, "learning_rate": 1.0165799476908827e-05, "loss": 0.0247, "step": 128600 }, { "epoch": 2.5700411654210464, "grad_norm": 0.16128531098365784, "learning_rate": 1.0159554386339999e-05, "loss": 0.027, "step": 128610 }, { "epoch": 2.5702409975620477, "grad_norm": 0.16216938197612762, "learning_rate": 1.0153310997681847e-05, "loss": 0.0214, "step": 128620 }, { "epoch": 2.5704408297030494, "grad_norm": 0.12434831261634827, "learning_rate": 1.0147069311201035e-05, "loss": 0.0258, "step": 128630 }, { "epoch": 2.570640661844051, "grad_norm": 0.09683673828840256, "learning_rate": 1.0140829327164208e-05, "loss": 0.0238, "step": 128640 }, { "epoch": 2.5708404939850524, "grad_norm": 0.1336909383535385, "learning_rate": 1.0134591045837932e-05, "loss": 0.0244, "step": 128650 }, { "epoch": 2.571040326126054, "grad_norm": 0.14553214609622955, "learning_rate": 1.0128354467488694e-05, "loss": 0.015, "step": 128660 }, { "epoch": 2.571240158267056, "grad_norm": 0.2047051042318344, "learning_rate": 1.0122119592382901e-05, "loss": 0.0255, "step": 128670 }, { "epoch": 2.571439990408057, "grad_norm": 0.1379736363887787, "learning_rate": 1.0115886420786902e-05, "loss": 0.0267, "step": 128680 }, { "epoch": 2.571639822549059, "grad_norm": 0.13574492931365967, "learning_rate": 1.0109654952966979e-05, "loss": 0.0277, "step": 128690 }, { "epoch": 2.57183965469006, "grad_norm": 0.10125850141048431, "learning_rate": 1.0103425189189292e-05, "loss": 0.0238, "step": 128700 }, { "epoch": 2.572039486831062, "grad_norm": 0.11885161697864532, "learning_rate": 1.0097197129719998e-05, "loss": 0.0228, "step": 128710 }, { "epoch": 2.5722393189720636, "grad_norm": 0.10573547333478928, "learning_rate": 1.0090970774825154e-05, "loss": 0.02, "step": 128720 }, { "epoch": 2.572439151113065, "grad_norm": 0.12348013371229172, "learning_rate": 1.0084746124770705e-05, "loss": 0.025, "step": 128730 }, { "epoch": 2.5726389832540666, "grad_norm": 0.15352095663547516, "learning_rate": 1.0078523179822569e-05, "loss": 0.0201, "step": 128740 }, { "epoch": 2.572838815395068, "grad_norm": 0.13768936693668365, "learning_rate": 1.0072301940246609e-05, "loss": 0.0282, "step": 128750 }, { "epoch": 2.5730386475360696, "grad_norm": 0.11164135485887527, "learning_rate": 1.0066082406308546e-05, "loss": 0.0277, "step": 128760 }, { "epoch": 2.5732384796770713, "grad_norm": 0.11326483637094498, "learning_rate": 1.005986457827408e-05, "loss": 0.0267, "step": 128770 }, { "epoch": 2.573438311818073, "grad_norm": 0.11928243935108185, "learning_rate": 1.0053648456408833e-05, "loss": 0.0195, "step": 128780 }, { "epoch": 2.5736381439590743, "grad_norm": 0.10777338594198227, "learning_rate": 1.0047434040978332e-05, "loss": 0.0281, "step": 128790 }, { "epoch": 2.573837976100076, "grad_norm": 0.06264431774616241, "learning_rate": 1.0041221332248057e-05, "loss": 0.0187, "step": 128800 }, { "epoch": 2.5740378082410773, "grad_norm": 0.10838737338781357, "learning_rate": 1.0035010330483401e-05, "loss": 0.0215, "step": 128810 }, { "epoch": 2.574237640382079, "grad_norm": 0.12695324420928955, "learning_rate": 1.0028801035949697e-05, "loss": 0.0329, "step": 128820 }, { "epoch": 2.5744374725230807, "grad_norm": 0.11123941093683243, "learning_rate": 1.002259344891216e-05, "loss": 0.0199, "step": 128830 }, { "epoch": 2.574637304664082, "grad_norm": 0.11313573271036148, "learning_rate": 1.0016387569636004e-05, "loss": 0.0214, "step": 128840 }, { "epoch": 2.5748371368050837, "grad_norm": 0.180621936917305, "learning_rate": 1.001018339838633e-05, "loss": 0.0281, "step": 128850 }, { "epoch": 2.575036968946085, "grad_norm": 0.252546489238739, "learning_rate": 1.0003980935428142e-05, "loss": 0.0331, "step": 128860 }, { "epoch": 2.5752368010870867, "grad_norm": 0.10049190372228622, "learning_rate": 9.997780181026417e-06, "loss": 0.0224, "step": 128870 }, { "epoch": 2.5754366332280885, "grad_norm": 0.16149787604808807, "learning_rate": 9.991581135446032e-06, "loss": 0.0266, "step": 128880 }, { "epoch": 2.57563646536909, "grad_norm": 0.16163720190525055, "learning_rate": 9.98538379895181e-06, "loss": 0.0238, "step": 128890 }, { "epoch": 2.5758362975100915, "grad_norm": 0.10128585994243622, "learning_rate": 9.979188171808478e-06, "loss": 0.0286, "step": 128900 }, { "epoch": 2.576036129651093, "grad_norm": 0.12600062787532806, "learning_rate": 9.972994254280705e-06, "loss": 0.0244, "step": 128910 }, { "epoch": 2.5762359617920945, "grad_norm": 0.15826644003391266, "learning_rate": 9.966802046633089e-06, "loss": 0.0251, "step": 128920 }, { "epoch": 2.576435793933096, "grad_norm": 0.11171042919158936, "learning_rate": 9.960611549130144e-06, "loss": 0.0259, "step": 128930 }, { "epoch": 2.576635626074098, "grad_norm": 0.1236673966050148, "learning_rate": 9.95442276203632e-06, "loss": 0.0264, "step": 128940 }, { "epoch": 2.576835458215099, "grad_norm": 0.12388401478528976, "learning_rate": 9.948235685616003e-06, "loss": 0.023, "step": 128950 }, { "epoch": 2.577035290356101, "grad_norm": 0.09006921947002411, "learning_rate": 9.942050320133467e-06, "loss": 0.0249, "step": 128960 }, { "epoch": 2.577235122497102, "grad_norm": 0.1235809400677681, "learning_rate": 9.935866665852938e-06, "loss": 0.0209, "step": 128970 }, { "epoch": 2.577434954638104, "grad_norm": 0.12018197029829025, "learning_rate": 9.929684723038612e-06, "loss": 0.0265, "step": 128980 }, { "epoch": 2.5776347867791056, "grad_norm": 0.118344746530056, "learning_rate": 9.923504491954534e-06, "loss": 0.0214, "step": 128990 }, { "epoch": 2.5778346189201073, "grad_norm": 0.13968360424041748, "learning_rate": 9.917325972864715e-06, "loss": 0.0297, "step": 129000 }, { "epoch": 2.5780344510611086, "grad_norm": 0.13557203114032745, "learning_rate": 9.911149166033084e-06, "loss": 0.0291, "step": 129010 }, { "epoch": 2.5782342832021103, "grad_norm": 0.144208624958992, "learning_rate": 9.904974071723545e-06, "loss": 0.025, "step": 129020 }, { "epoch": 2.5784341153431116, "grad_norm": 0.09616672992706299, "learning_rate": 9.898800690199844e-06, "loss": 0.0214, "step": 129030 }, { "epoch": 2.5786339474841133, "grad_norm": 0.1512918919324875, "learning_rate": 9.892629021725702e-06, "loss": 0.0242, "step": 129040 }, { "epoch": 2.578833779625115, "grad_norm": 0.11978308856487274, "learning_rate": 9.886459066564774e-06, "loss": 0.0239, "step": 129050 }, { "epoch": 2.5790336117661163, "grad_norm": 0.13587477803230286, "learning_rate": 9.880290824980626e-06, "loss": 0.0312, "step": 129060 }, { "epoch": 2.579233443907118, "grad_norm": 0.11466824263334274, "learning_rate": 9.874124297236748e-06, "loss": 0.0259, "step": 129070 }, { "epoch": 2.5794332760481193, "grad_norm": 0.12935204803943634, "learning_rate": 9.867959483596567e-06, "loss": 0.0171, "step": 129080 }, { "epoch": 2.579633108189121, "grad_norm": 0.09930059313774109, "learning_rate": 9.861796384323446e-06, "loss": 0.0239, "step": 129090 }, { "epoch": 2.579832940330123, "grad_norm": 0.1088133156299591, "learning_rate": 9.855634999680618e-06, "loss": 0.0285, "step": 129100 }, { "epoch": 2.5800327724711245, "grad_norm": 0.09588418900966644, "learning_rate": 9.84947532993133e-06, "loss": 0.0198, "step": 129110 }, { "epoch": 2.580232604612126, "grad_norm": 0.11904184520244598, "learning_rate": 9.843317375338713e-06, "loss": 0.0235, "step": 129120 }, { "epoch": 2.5804324367531275, "grad_norm": 0.14144501090049744, "learning_rate": 9.83716113616579e-06, "loss": 0.0271, "step": 129130 }, { "epoch": 2.580632268894129, "grad_norm": 0.12171021103858948, "learning_rate": 9.831006612675552e-06, "loss": 0.0252, "step": 129140 }, { "epoch": 2.5808321010351305, "grad_norm": 0.12031809240579605, "learning_rate": 9.824853805130945e-06, "loss": 0.0263, "step": 129150 }, { "epoch": 2.581031933176132, "grad_norm": 0.15562647581100464, "learning_rate": 9.818702713794765e-06, "loss": 0.0252, "step": 129160 }, { "epoch": 2.5812317653171335, "grad_norm": 0.16269166767597198, "learning_rate": 9.812553338929786e-06, "loss": 0.0212, "step": 129170 }, { "epoch": 2.581431597458135, "grad_norm": 0.14254723489284515, "learning_rate": 9.806405680798708e-06, "loss": 0.0246, "step": 129180 }, { "epoch": 2.5816314295991365, "grad_norm": 0.11257847398519516, "learning_rate": 9.800259739664141e-06, "loss": 0.0348, "step": 129190 }, { "epoch": 2.581831261740138, "grad_norm": 0.10372044146060944, "learning_rate": 9.794115515788627e-06, "loss": 0.024, "step": 129200 }, { "epoch": 2.58203109388114, "grad_norm": 0.12308565527200699, "learning_rate": 9.787973009434637e-06, "loss": 0.0309, "step": 129210 }, { "epoch": 2.5822309260221417, "grad_norm": 0.0968838557600975, "learning_rate": 9.781832220864579e-06, "loss": 0.0284, "step": 129220 }, { "epoch": 2.582430758163143, "grad_norm": 0.07192704826593399, "learning_rate": 9.775693150340743e-06, "loss": 0.0206, "step": 129230 }, { "epoch": 2.5826305903041447, "grad_norm": 0.1166667714715004, "learning_rate": 9.769555798125413e-06, "loss": 0.0287, "step": 129240 }, { "epoch": 2.582830422445146, "grad_norm": 0.1213764026761055, "learning_rate": 9.763420164480763e-06, "loss": 0.0348, "step": 129250 }, { "epoch": 2.5830302545861477, "grad_norm": 0.17521429061889648, "learning_rate": 9.75728624966888e-06, "loss": 0.0256, "step": 129260 }, { "epoch": 2.5832300867271494, "grad_norm": 0.11771514266729355, "learning_rate": 9.751154053951794e-06, "loss": 0.02, "step": 129270 }, { "epoch": 2.5834299188681507, "grad_norm": 0.08871682733297348, "learning_rate": 9.745023577591472e-06, "loss": 0.0266, "step": 129280 }, { "epoch": 2.5836297510091524, "grad_norm": 0.08918258547782898, "learning_rate": 9.738894820849793e-06, "loss": 0.0179, "step": 129290 }, { "epoch": 2.5838295831501537, "grad_norm": 0.13166029751300812, "learning_rate": 9.732767783988562e-06, "loss": 0.0221, "step": 129300 }, { "epoch": 2.5840294152911554, "grad_norm": 0.13734278082847595, "learning_rate": 9.726642467269514e-06, "loss": 0.0263, "step": 129310 }, { "epoch": 2.584229247432157, "grad_norm": 0.1470661461353302, "learning_rate": 9.72051887095432e-06, "loss": 0.0276, "step": 129320 }, { "epoch": 2.584429079573159, "grad_norm": 0.11696676909923553, "learning_rate": 9.714396995304564e-06, "loss": 0.0257, "step": 129330 }, { "epoch": 2.58462891171416, "grad_norm": 0.12655487656593323, "learning_rate": 9.708276840581764e-06, "loss": 0.0241, "step": 129340 }, { "epoch": 2.584828743855162, "grad_norm": 0.12372832000255585, "learning_rate": 9.702158407047368e-06, "loss": 0.0323, "step": 129350 }, { "epoch": 2.585028575996163, "grad_norm": 0.14737333357334137, "learning_rate": 9.696041694962726e-06, "loss": 0.027, "step": 129360 }, { "epoch": 2.585228408137165, "grad_norm": 0.10653900355100632, "learning_rate": 9.689926704589125e-06, "loss": 0.0269, "step": 129370 }, { "epoch": 2.5854282402781665, "grad_norm": 0.1455444097518921, "learning_rate": 9.683813436187833e-06, "loss": 0.026, "step": 129380 }, { "epoch": 2.585628072419168, "grad_norm": 0.0846615880727768, "learning_rate": 9.677701890019952e-06, "loss": 0.0236, "step": 129390 }, { "epoch": 2.5858279045601695, "grad_norm": 0.17051884531974792, "learning_rate": 9.671592066346575e-06, "loss": 0.0265, "step": 129400 }, { "epoch": 2.586027736701171, "grad_norm": 0.07944539189338684, "learning_rate": 9.66548396542868e-06, "loss": 0.0204, "step": 129410 }, { "epoch": 2.5862275688421725, "grad_norm": 0.1115860566496849, "learning_rate": 9.659377587527241e-06, "loss": 0.0184, "step": 129420 }, { "epoch": 2.5864274009831743, "grad_norm": 0.11311567574739456, "learning_rate": 9.653272932903074e-06, "loss": 0.0303, "step": 129430 }, { "epoch": 2.5866272331241755, "grad_norm": 0.12141703814268112, "learning_rate": 9.647170001816968e-06, "loss": 0.0258, "step": 129440 }, { "epoch": 2.5868270652651773, "grad_norm": 0.10615269094705582, "learning_rate": 9.64106879452963e-06, "loss": 0.0246, "step": 129450 }, { "epoch": 2.587026897406179, "grad_norm": 0.1405125856399536, "learning_rate": 9.634969311301683e-06, "loss": 0.0198, "step": 129460 }, { "epoch": 2.5872267295471802, "grad_norm": 0.11256370693445206, "learning_rate": 9.628871552393703e-06, "loss": 0.0215, "step": 129470 }, { "epoch": 2.587426561688182, "grad_norm": 0.09641627967357635, "learning_rate": 9.622775518066163e-06, "loss": 0.0259, "step": 129480 }, { "epoch": 2.5876263938291837, "grad_norm": 0.12090378999710083, "learning_rate": 9.616681208579492e-06, "loss": 0.0282, "step": 129490 }, { "epoch": 2.587826225970185, "grad_norm": 0.13594169914722443, "learning_rate": 9.610588624193984e-06, "loss": 0.0244, "step": 129500 }, { "epoch": 2.5880260581111867, "grad_norm": 0.18684083223342896, "learning_rate": 9.604497765169945e-06, "loss": 0.0323, "step": 129510 }, { "epoch": 2.588225890252188, "grad_norm": 0.12153401225805283, "learning_rate": 9.598408631767568e-06, "loss": 0.0273, "step": 129520 }, { "epoch": 2.5884257223931897, "grad_norm": 0.11990039795637131, "learning_rate": 9.592321224246942e-06, "loss": 0.0283, "step": 129530 }, { "epoch": 2.5886255545341914, "grad_norm": 0.08386451750993729, "learning_rate": 9.586235542868121e-06, "loss": 0.0247, "step": 129540 }, { "epoch": 2.5888253866751927, "grad_norm": 0.11935889720916748, "learning_rate": 9.580151587891074e-06, "loss": 0.025, "step": 129550 }, { "epoch": 2.5890252188161944, "grad_norm": 0.11293110251426697, "learning_rate": 9.574069359575699e-06, "loss": 0.0178, "step": 129560 }, { "epoch": 2.5892250509571957, "grad_norm": 0.1532883495092392, "learning_rate": 9.567988858181815e-06, "loss": 0.0308, "step": 129570 }, { "epoch": 2.5894248830981974, "grad_norm": 0.12517550587654114, "learning_rate": 9.561910083969173e-06, "loss": 0.024, "step": 129580 }, { "epoch": 2.589624715239199, "grad_norm": 0.12257274240255356, "learning_rate": 9.555833037197443e-06, "loss": 0.0268, "step": 129590 }, { "epoch": 2.589824547380201, "grad_norm": 0.1345038264989853, "learning_rate": 9.549757718126229e-06, "loss": 0.0246, "step": 129600 }, { "epoch": 2.590024379521202, "grad_norm": 0.12165601551532745, "learning_rate": 9.543684127015056e-06, "loss": 0.0291, "step": 129610 }, { "epoch": 2.590224211662204, "grad_norm": 0.1111968606710434, "learning_rate": 9.537612264123385e-06, "loss": 0.0238, "step": 129620 }, { "epoch": 2.590424043803205, "grad_norm": 0.0936015248298645, "learning_rate": 9.53154212971058e-06, "loss": 0.0244, "step": 129630 }, { "epoch": 2.590623875944207, "grad_norm": 0.12128342688083649, "learning_rate": 9.525473724035933e-06, "loss": 0.0178, "step": 129640 }, { "epoch": 2.5908237080852086, "grad_norm": 0.16133518517017365, "learning_rate": 9.519407047358725e-06, "loss": 0.0288, "step": 129650 }, { "epoch": 2.59102354022621, "grad_norm": 0.07994309812784195, "learning_rate": 9.513342099938066e-06, "loss": 0.0375, "step": 129660 }, { "epoch": 2.5912233723672116, "grad_norm": 0.15322183072566986, "learning_rate": 9.507278882033061e-06, "loss": 0.025, "step": 129670 }, { "epoch": 2.591423204508213, "grad_norm": 0.13611352443695068, "learning_rate": 9.501217393902712e-06, "loss": 0.0348, "step": 129680 }, { "epoch": 2.5916230366492146, "grad_norm": 0.09141965210437775, "learning_rate": 9.495157635805957e-06, "loss": 0.0311, "step": 129690 }, { "epoch": 2.5918228687902163, "grad_norm": 0.11622761189937592, "learning_rate": 9.489099608001657e-06, "loss": 0.028, "step": 129700 }, { "epoch": 2.592022700931218, "grad_norm": 0.13922575116157532, "learning_rate": 9.483043310748602e-06, "loss": 0.0259, "step": 129710 }, { "epoch": 2.5922225330722193, "grad_norm": 0.08763714879751205, "learning_rate": 9.476988744305521e-06, "loss": 0.0264, "step": 129720 }, { "epoch": 2.592422365213221, "grad_norm": 0.11574714630842209, "learning_rate": 9.470935908931005e-06, "loss": 0.0245, "step": 129730 }, { "epoch": 2.5926221973542223, "grad_norm": 0.12303754687309265, "learning_rate": 9.464884804883672e-06, "loss": 0.0299, "step": 129740 }, { "epoch": 2.592822029495224, "grad_norm": 0.09347094595432281, "learning_rate": 9.458835432422003e-06, "loss": 0.0225, "step": 129750 }, { "epoch": 2.5930218616362257, "grad_norm": 0.13134384155273438, "learning_rate": 9.452787791804397e-06, "loss": 0.0211, "step": 129760 }, { "epoch": 2.593221693777227, "grad_norm": 0.11038067936897278, "learning_rate": 9.446741883289196e-06, "loss": 0.0262, "step": 129770 }, { "epoch": 2.5934215259182287, "grad_norm": 0.11003559082746506, "learning_rate": 9.44069770713471e-06, "loss": 0.0244, "step": 129780 }, { "epoch": 2.59362135805923, "grad_norm": 0.13455113768577576, "learning_rate": 9.43465526359909e-06, "loss": 0.0292, "step": 129790 }, { "epoch": 2.5938211902002317, "grad_norm": 0.11919404566287994, "learning_rate": 9.428614552940479e-06, "loss": 0.0189, "step": 129800 }, { "epoch": 2.5940210223412334, "grad_norm": 0.13702943921089172, "learning_rate": 9.422575575416925e-06, "loss": 0.0237, "step": 129810 }, { "epoch": 2.594220854482235, "grad_norm": 0.18787376582622528, "learning_rate": 9.416538331286401e-06, "loss": 0.0241, "step": 129820 }, { "epoch": 2.5944206866232364, "grad_norm": 0.12279333174228668, "learning_rate": 9.4105028208068e-06, "loss": 0.0188, "step": 129830 }, { "epoch": 2.594620518764238, "grad_norm": 0.1272834837436676, "learning_rate": 9.404469044235953e-06, "loss": 0.0246, "step": 129840 }, { "epoch": 2.5948203509052394, "grad_norm": 0.11746105551719666, "learning_rate": 9.398437001831616e-06, "loss": 0.0248, "step": 129850 }, { "epoch": 2.595020183046241, "grad_norm": 0.10058258473873138, "learning_rate": 9.392406693851463e-06, "loss": 0.0234, "step": 129860 }, { "epoch": 2.595220015187243, "grad_norm": 0.20977340638637543, "learning_rate": 9.38637812055309e-06, "loss": 0.0294, "step": 129870 }, { "epoch": 2.595419847328244, "grad_norm": 0.08496590703725815, "learning_rate": 9.380351282194044e-06, "loss": 0.0269, "step": 129880 }, { "epoch": 2.595619679469246, "grad_norm": 0.2582116425037384, "learning_rate": 9.374326179031778e-06, "loss": 0.0245, "step": 129890 }, { "epoch": 2.595819511610247, "grad_norm": 0.08770278841257095, "learning_rate": 9.368302811323655e-06, "loss": 0.0249, "step": 129900 }, { "epoch": 2.596019343751249, "grad_norm": 0.4168310761451721, "learning_rate": 9.362281179326976e-06, "loss": 0.0231, "step": 129910 }, { "epoch": 2.5962191758922506, "grad_norm": 0.13819094002246857, "learning_rate": 9.356261283299023e-06, "loss": 0.0242, "step": 129920 }, { "epoch": 2.5964190080332523, "grad_norm": 0.16451437771320343, "learning_rate": 9.350243123496904e-06, "loss": 0.0266, "step": 129930 }, { "epoch": 2.5966188401742536, "grad_norm": 0.10240976512432098, "learning_rate": 9.344226700177728e-06, "loss": 0.0214, "step": 129940 }, { "epoch": 2.5968186723152553, "grad_norm": 0.11381742358207703, "learning_rate": 9.338212013598497e-06, "loss": 0.0255, "step": 129950 }, { "epoch": 2.5970185044562566, "grad_norm": 0.12672270834445953, "learning_rate": 9.332199064016156e-06, "loss": 0.0245, "step": 129960 }, { "epoch": 2.5972183365972583, "grad_norm": 0.12872347235679626, "learning_rate": 9.326187851687556e-06, "loss": 0.0275, "step": 129970 }, { "epoch": 2.59741816873826, "grad_norm": 0.10358669608831406, "learning_rate": 9.320178376869493e-06, "loss": 0.0344, "step": 129980 }, { "epoch": 2.5976180008792613, "grad_norm": 0.11821258813142776, "learning_rate": 9.314170639818693e-06, "loss": 0.0352, "step": 129990 }, { "epoch": 2.597817833020263, "grad_norm": 0.1457095742225647, "learning_rate": 9.30816464079175e-06, "loss": 0.0189, "step": 130000 }, { "epoch": 2.5980176651612643, "grad_norm": 0.15045815706253052, "learning_rate": 9.302160380045277e-06, "loss": 0.0301, "step": 130010 }, { "epoch": 2.598217497302266, "grad_norm": 0.12999090552330017, "learning_rate": 9.296157857835763e-06, "loss": 0.0273, "step": 130020 }, { "epoch": 2.5984173294432678, "grad_norm": 0.12174414843320847, "learning_rate": 9.290157074419592e-06, "loss": 0.02, "step": 130030 }, { "epoch": 2.5986171615842695, "grad_norm": 0.12628747522830963, "learning_rate": 9.284158030053114e-06, "loss": 0.0258, "step": 130040 }, { "epoch": 2.5988169937252708, "grad_norm": 0.1155647560954094, "learning_rate": 9.278160724992634e-06, "loss": 0.0312, "step": 130050 }, { "epoch": 2.5990168258662725, "grad_norm": 0.1049233227968216, "learning_rate": 9.272165159494307e-06, "loss": 0.0253, "step": 130060 }, { "epoch": 2.5992166580072738, "grad_norm": 0.12000445276498795, "learning_rate": 9.266171333814262e-06, "loss": 0.0224, "step": 130070 }, { "epoch": 2.5994164901482755, "grad_norm": 0.08637993037700653, "learning_rate": 9.26017924820855e-06, "loss": 0.0207, "step": 130080 }, { "epoch": 2.599616322289277, "grad_norm": 0.11995294690132141, "learning_rate": 9.254188902933136e-06, "loss": 0.0285, "step": 130090 }, { "epoch": 2.5998161544302785, "grad_norm": 0.10870083421468735, "learning_rate": 9.248200298243925e-06, "loss": 0.0242, "step": 130100 }, { "epoch": 2.60001598657128, "grad_norm": 0.10909707844257355, "learning_rate": 9.242213434396729e-06, "loss": 0.0214, "step": 130110 }, { "epoch": 2.6002158187122815, "grad_norm": 0.10285324603319168, "learning_rate": 9.236228311647316e-06, "loss": 0.025, "step": 130120 }, { "epoch": 2.600415650853283, "grad_norm": 0.09052146971225739, "learning_rate": 9.23024493025132e-06, "loss": 0.0214, "step": 130130 }, { "epoch": 2.600615482994285, "grad_norm": 0.12494909018278122, "learning_rate": 9.224263290464375e-06, "loss": 0.023, "step": 130140 }, { "epoch": 2.6008153151352866, "grad_norm": 0.08666809648275375, "learning_rate": 9.218283392542005e-06, "loss": 0.0258, "step": 130150 }, { "epoch": 2.601015147276288, "grad_norm": 0.14308136701583862, "learning_rate": 9.212305236739637e-06, "loss": 0.0269, "step": 130160 }, { "epoch": 2.6012149794172896, "grad_norm": 0.09343526512384415, "learning_rate": 9.206328823312666e-06, "loss": 0.0233, "step": 130170 }, { "epoch": 2.601414811558291, "grad_norm": 0.11069530248641968, "learning_rate": 9.200354152516371e-06, "loss": 0.0253, "step": 130180 }, { "epoch": 2.6016146436992926, "grad_norm": 0.08012134581804276, "learning_rate": 9.194381224606018e-06, "loss": 0.0258, "step": 130190 }, { "epoch": 2.6018144758402943, "grad_norm": 0.11735623329877853, "learning_rate": 9.188410039836731e-06, "loss": 0.0231, "step": 130200 }, { "epoch": 2.6020143079812956, "grad_norm": 0.13487586379051208, "learning_rate": 9.182440598463587e-06, "loss": 0.0241, "step": 130210 }, { "epoch": 2.6022141401222973, "grad_norm": 0.14010562002658844, "learning_rate": 9.1764729007416e-06, "loss": 0.0264, "step": 130220 }, { "epoch": 2.6024139722632986, "grad_norm": 0.12157661467790604, "learning_rate": 9.170506946925694e-06, "loss": 0.0252, "step": 130230 }, { "epoch": 2.6026138044043003, "grad_norm": 0.08572432398796082, "learning_rate": 9.164542737270727e-06, "loss": 0.0237, "step": 130240 }, { "epoch": 2.602813636545302, "grad_norm": 0.09929172694683075, "learning_rate": 9.158580272031474e-06, "loss": 0.0281, "step": 130250 }, { "epoch": 2.6030134686863033, "grad_norm": 0.12655048072338104, "learning_rate": 9.152619551462643e-06, "loss": 0.0342, "step": 130260 }, { "epoch": 2.603213300827305, "grad_norm": 0.12029159069061279, "learning_rate": 9.146660575818871e-06, "loss": 0.0261, "step": 130270 }, { "epoch": 2.603413132968307, "grad_norm": 0.11144412308931351, "learning_rate": 9.140703345354706e-06, "loss": 0.0251, "step": 130280 }, { "epoch": 2.603612965109308, "grad_norm": 0.12319003790616989, "learning_rate": 9.134747860324644e-06, "loss": 0.028, "step": 130290 }, { "epoch": 2.60381279725031, "grad_norm": 0.11809088289737701, "learning_rate": 9.12879412098307e-06, "loss": 0.027, "step": 130300 }, { "epoch": 2.6040126293913115, "grad_norm": 0.12440522760152817, "learning_rate": 9.122842127584319e-06, "loss": 0.0196, "step": 130310 }, { "epoch": 2.604212461532313, "grad_norm": 0.1378161758184433, "learning_rate": 9.116891880382684e-06, "loss": 0.0249, "step": 130320 }, { "epoch": 2.6044122936733145, "grad_norm": 0.18117426335811615, "learning_rate": 9.110943379632304e-06, "loss": 0.0245, "step": 130330 }, { "epoch": 2.604612125814316, "grad_norm": 0.11540531367063522, "learning_rate": 9.104996625587314e-06, "loss": 0.0267, "step": 130340 }, { "epoch": 2.6048119579553175, "grad_norm": 0.1541072577238083, "learning_rate": 9.099051618501737e-06, "loss": 0.0243, "step": 130350 }, { "epoch": 2.6050117900963192, "grad_norm": 0.11903714388608932, "learning_rate": 9.093108358629538e-06, "loss": 0.017, "step": 130360 }, { "epoch": 2.6052116222373205, "grad_norm": 0.10063296556472778, "learning_rate": 9.087166846224604e-06, "loss": 0.0192, "step": 130370 }, { "epoch": 2.6054114543783222, "grad_norm": 0.12581785023212433, "learning_rate": 9.081227081540744e-06, "loss": 0.0353, "step": 130380 }, { "epoch": 2.6056112865193235, "grad_norm": 0.11034294962882996, "learning_rate": 9.0752890648317e-06, "loss": 0.0187, "step": 130390 }, { "epoch": 2.605811118660325, "grad_norm": 0.11632072925567627, "learning_rate": 9.069352796351106e-06, "loss": 0.0181, "step": 130400 }, { "epoch": 2.606010950801327, "grad_norm": 0.14113068580627441, "learning_rate": 9.063418276352581e-06, "loss": 0.0296, "step": 130410 }, { "epoch": 2.6062107829423287, "grad_norm": 0.13172051310539246, "learning_rate": 9.057485505089636e-06, "loss": 0.0311, "step": 130420 }, { "epoch": 2.60641061508333, "grad_norm": 0.07908178865909576, "learning_rate": 9.051554482815682e-06, "loss": 0.0265, "step": 130430 }, { "epoch": 2.6066104472243317, "grad_norm": 0.15464073419570923, "learning_rate": 9.045625209784081e-06, "loss": 0.0302, "step": 130440 }, { "epoch": 2.606810279365333, "grad_norm": 0.09304531663656235, "learning_rate": 9.039697686248167e-06, "loss": 0.0195, "step": 130450 }, { "epoch": 2.6070101115063347, "grad_norm": 0.12640812993049622, "learning_rate": 9.033771912461103e-06, "loss": 0.0245, "step": 130460 }, { "epoch": 2.6072099436473364, "grad_norm": 0.12921853363513947, "learning_rate": 9.02784788867605e-06, "loss": 0.0265, "step": 130470 }, { "epoch": 2.6074097757883377, "grad_norm": 0.12117451429367065, "learning_rate": 9.021925615146066e-06, "loss": 0.0231, "step": 130480 }, { "epoch": 2.6076096079293394, "grad_norm": 0.16230885684490204, "learning_rate": 9.016005092124142e-06, "loss": 0.0296, "step": 130490 }, { "epoch": 2.6078094400703407, "grad_norm": 0.09804736077785492, "learning_rate": 9.010086319863187e-06, "loss": 0.0258, "step": 130500 }, { "epoch": 2.6080092722113424, "grad_norm": 0.12849348783493042, "learning_rate": 9.004169298616044e-06, "loss": 0.0237, "step": 130510 }, { "epoch": 2.608209104352344, "grad_norm": 0.10375326126813889, "learning_rate": 8.99825402863549e-06, "loss": 0.0213, "step": 130520 }, { "epoch": 2.608408936493346, "grad_norm": 0.11085958033800125, "learning_rate": 8.992340510174175e-06, "loss": 0.017, "step": 130530 }, { "epoch": 2.608608768634347, "grad_norm": 0.12319402396678925, "learning_rate": 8.986428743484754e-06, "loss": 0.0184, "step": 130540 }, { "epoch": 2.608808600775349, "grad_norm": 0.11322477459907532, "learning_rate": 8.980518728819764e-06, "loss": 0.03, "step": 130550 }, { "epoch": 2.60900843291635, "grad_norm": 0.1412140280008316, "learning_rate": 8.974610466431648e-06, "loss": 0.0317, "step": 130560 }, { "epoch": 2.609208265057352, "grad_norm": 0.21726034581661224, "learning_rate": 8.968703956572804e-06, "loss": 0.0218, "step": 130570 }, { "epoch": 2.6094080971983535, "grad_norm": 0.09603434056043625, "learning_rate": 8.962799199495542e-06, "loss": 0.0225, "step": 130580 }, { "epoch": 2.609607929339355, "grad_norm": 0.18145915865898132, "learning_rate": 8.956896195452136e-06, "loss": 0.0342, "step": 130590 }, { "epoch": 2.6098077614803565, "grad_norm": 0.10088002681732178, "learning_rate": 8.950994944694707e-06, "loss": 0.0267, "step": 130600 }, { "epoch": 2.610007593621358, "grad_norm": 0.11250364035367966, "learning_rate": 8.945095447475371e-06, "loss": 0.0194, "step": 130610 }, { "epoch": 2.6102074257623595, "grad_norm": 0.1363106667995453, "learning_rate": 8.939197704046131e-06, "loss": 0.018, "step": 130620 }, { "epoch": 2.6104072579033613, "grad_norm": 0.10326872766017914, "learning_rate": 8.933301714658937e-06, "loss": 0.0211, "step": 130630 }, { "epoch": 2.610607090044363, "grad_norm": 0.15045900642871857, "learning_rate": 8.927407479565653e-06, "loss": 0.0254, "step": 130640 }, { "epoch": 2.6108069221853643, "grad_norm": 0.06893697381019592, "learning_rate": 8.921514999018065e-06, "loss": 0.0224, "step": 130650 }, { "epoch": 2.611006754326366, "grad_norm": 0.1088922768831253, "learning_rate": 8.915624273267913e-06, "loss": 0.0198, "step": 130660 }, { "epoch": 2.6112065864673673, "grad_norm": 0.17166952788829803, "learning_rate": 8.909735302566785e-06, "loss": 0.0347, "step": 130670 }, { "epoch": 2.611406418608369, "grad_norm": 0.10637915879487991, "learning_rate": 8.903848087166294e-06, "loss": 0.0173, "step": 130680 }, { "epoch": 2.6116062507493707, "grad_norm": 0.1276959478855133, "learning_rate": 8.89796262731793e-06, "loss": 0.0245, "step": 130690 }, { "epoch": 2.611806082890372, "grad_norm": 0.10095193982124329, "learning_rate": 8.892078923273084e-06, "loss": 0.021, "step": 130700 }, { "epoch": 2.6120059150313737, "grad_norm": 0.14357586205005646, "learning_rate": 8.886196975283096e-06, "loss": 0.029, "step": 130710 }, { "epoch": 2.612205747172375, "grad_norm": 0.13649269938468933, "learning_rate": 8.880316783599269e-06, "loss": 0.0216, "step": 130720 }, { "epoch": 2.6124055793133767, "grad_norm": 0.09130658209323883, "learning_rate": 8.874438348472758e-06, "loss": 0.0232, "step": 130730 }, { "epoch": 2.6126054114543784, "grad_norm": 0.13000883162021637, "learning_rate": 8.86856167015469e-06, "loss": 0.0274, "step": 130740 }, { "epoch": 2.61280524359538, "grad_norm": 0.1172436997294426, "learning_rate": 8.862686748896104e-06, "loss": 0.022, "step": 130750 }, { "epoch": 2.6130050757363814, "grad_norm": 0.09800839424133301, "learning_rate": 8.856813584947975e-06, "loss": 0.0226, "step": 130760 }, { "epoch": 2.613204907877383, "grad_norm": 0.11643615365028381, "learning_rate": 8.850942178561183e-06, "loss": 0.0232, "step": 130770 }, { "epoch": 2.6134047400183844, "grad_norm": 0.13091911375522614, "learning_rate": 8.845072529986548e-06, "loss": 0.027, "step": 130780 }, { "epoch": 2.613604572159386, "grad_norm": 0.11614585667848587, "learning_rate": 8.839204639474824e-06, "loss": 0.0292, "step": 130790 }, { "epoch": 2.613804404300388, "grad_norm": 0.11858098953962326, "learning_rate": 8.833338507276634e-06, "loss": 0.0346, "step": 130800 }, { "epoch": 2.614004236441389, "grad_norm": 0.1426122635602951, "learning_rate": 8.827474133642616e-06, "loss": 0.0272, "step": 130810 }, { "epoch": 2.614204068582391, "grad_norm": 0.12623876333236694, "learning_rate": 8.821611518823275e-06, "loss": 0.0243, "step": 130820 }, { "epoch": 2.614403900723392, "grad_norm": 0.13071545958518982, "learning_rate": 8.815750663069034e-06, "loss": 0.0225, "step": 130830 }, { "epoch": 2.614603732864394, "grad_norm": 0.1304154098033905, "learning_rate": 8.809891566630263e-06, "loss": 0.026, "step": 130840 }, { "epoch": 2.6148035650053956, "grad_norm": 0.0939553901553154, "learning_rate": 8.804034229757257e-06, "loss": 0.0249, "step": 130850 }, { "epoch": 2.6150033971463973, "grad_norm": 0.11678009480237961, "learning_rate": 8.79817865270023e-06, "loss": 0.0194, "step": 130860 }, { "epoch": 2.6152032292873986, "grad_norm": 0.12945730984210968, "learning_rate": 8.792324835709314e-06, "loss": 0.0218, "step": 130870 }, { "epoch": 2.6154030614284003, "grad_norm": 0.0974501371383667, "learning_rate": 8.786472779034589e-06, "loss": 0.0234, "step": 130880 }, { "epoch": 2.6156028935694016, "grad_norm": 0.14302295446395874, "learning_rate": 8.780622482926032e-06, "loss": 0.018, "step": 130890 }, { "epoch": 2.6158027257104033, "grad_norm": 0.11099299043416977, "learning_rate": 8.774773947633564e-06, "loss": 0.0255, "step": 130900 }, { "epoch": 2.616002557851405, "grad_norm": 0.10075101256370544, "learning_rate": 8.768927173407015e-06, "loss": 0.027, "step": 130910 }, { "epoch": 2.6162023899924063, "grad_norm": 0.10623185336589813, "learning_rate": 8.763082160496167e-06, "loss": 0.0292, "step": 130920 }, { "epoch": 2.616402222133408, "grad_norm": 0.1304062157869339, "learning_rate": 8.757238909150678e-06, "loss": 0.0263, "step": 130930 }, { "epoch": 2.6166020542744093, "grad_norm": 0.10413138568401337, "learning_rate": 8.751397419620172e-06, "loss": 0.0219, "step": 130940 }, { "epoch": 2.616801886415411, "grad_norm": 0.11110617220401764, "learning_rate": 8.74555769215421e-06, "loss": 0.0237, "step": 130950 }, { "epoch": 2.6170017185564127, "grad_norm": 0.11680249124765396, "learning_rate": 8.739719727002233e-06, "loss": 0.0176, "step": 130960 }, { "epoch": 2.6172015506974144, "grad_norm": 0.1480710804462433, "learning_rate": 8.733883524413627e-06, "loss": 0.029, "step": 130970 }, { "epoch": 2.6174013828384157, "grad_norm": 0.15425178408622742, "learning_rate": 8.728049084637696e-06, "loss": 0.0279, "step": 130980 }, { "epoch": 2.6176012149794174, "grad_norm": 0.11678708344697952, "learning_rate": 8.722216407923711e-06, "loss": 0.0202, "step": 130990 }, { "epoch": 2.6178010471204187, "grad_norm": 0.20441493391990662, "learning_rate": 8.716385494520802e-06, "loss": 0.0293, "step": 131000 }, { "epoch": 2.6180008792614204, "grad_norm": 0.06931600719690323, "learning_rate": 8.710556344678066e-06, "loss": 0.023, "step": 131010 }, { "epoch": 2.618200711402422, "grad_norm": 0.11699032038450241, "learning_rate": 8.704728958644508e-06, "loss": 0.033, "step": 131020 }, { "epoch": 2.6184005435434234, "grad_norm": 0.1285015344619751, "learning_rate": 8.698903336669068e-06, "loss": 0.0273, "step": 131030 }, { "epoch": 2.618600375684425, "grad_norm": 0.13361501693725586, "learning_rate": 8.693079479000604e-06, "loss": 0.0241, "step": 131040 }, { "epoch": 2.6188002078254264, "grad_norm": 0.11247311532497406, "learning_rate": 8.687257385887904e-06, "loss": 0.0319, "step": 131050 }, { "epoch": 2.619000039966428, "grad_norm": 0.15164412558078766, "learning_rate": 8.68143705757969e-06, "loss": 0.0268, "step": 131060 }, { "epoch": 2.61919987210743, "grad_norm": 0.11559826880693436, "learning_rate": 8.67561849432455e-06, "loss": 0.0229, "step": 131070 }, { "epoch": 2.619399704248431, "grad_norm": 0.11681020259857178, "learning_rate": 8.669801696371087e-06, "loss": 0.0308, "step": 131080 }, { "epoch": 2.619599536389433, "grad_norm": 0.09280046820640564, "learning_rate": 8.663986663967783e-06, "loss": 0.0276, "step": 131090 }, { "epoch": 2.619799368530434, "grad_norm": 0.10461858659982681, "learning_rate": 8.658173397363017e-06, "loss": 0.0304, "step": 131100 }, { "epoch": 2.619999200671436, "grad_norm": 0.10363443195819855, "learning_rate": 8.652361896805139e-06, "loss": 0.0249, "step": 131110 }, { "epoch": 2.6201990328124376, "grad_norm": 0.1345391571521759, "learning_rate": 8.646552162542399e-06, "loss": 0.0232, "step": 131120 }, { "epoch": 2.6203988649534393, "grad_norm": 0.11677778512239456, "learning_rate": 8.640744194822986e-06, "loss": 0.0278, "step": 131130 }, { "epoch": 2.6205986970944406, "grad_norm": 0.12935547530651093, "learning_rate": 8.634937993894998e-06, "loss": 0.0276, "step": 131140 }, { "epoch": 2.6207985292354423, "grad_norm": 0.10385313630104065, "learning_rate": 8.629133560006469e-06, "loss": 0.0338, "step": 131150 }, { "epoch": 2.6209983613764436, "grad_norm": 0.1441717892885208, "learning_rate": 8.623330893405356e-06, "loss": 0.0231, "step": 131160 }, { "epoch": 2.6211981935174453, "grad_norm": 0.07114061713218689, "learning_rate": 8.617529994339535e-06, "loss": 0.0213, "step": 131170 }, { "epoch": 2.621398025658447, "grad_norm": 0.11852330714464188, "learning_rate": 8.611730863056811e-06, "loss": 0.0235, "step": 131180 }, { "epoch": 2.6215978577994483, "grad_norm": 0.11814845353364944, "learning_rate": 8.605933499804925e-06, "loss": 0.0196, "step": 131190 }, { "epoch": 2.62179768994045, "grad_norm": 0.12233298271894455, "learning_rate": 8.600137904831501e-06, "loss": 0.0185, "step": 131200 }, { "epoch": 2.6219975220814513, "grad_norm": 0.10371912270784378, "learning_rate": 8.59434407838412e-06, "loss": 0.027, "step": 131210 }, { "epoch": 2.622197354222453, "grad_norm": 0.16914363205432892, "learning_rate": 8.588552020710321e-06, "loss": 0.0249, "step": 131220 }, { "epoch": 2.6223971863634548, "grad_norm": 0.15296731889247894, "learning_rate": 8.582761732057493e-06, "loss": 0.0258, "step": 131230 }, { "epoch": 2.6225970185044565, "grad_norm": 0.1164000853896141, "learning_rate": 8.576973212672996e-06, "loss": 0.0232, "step": 131240 }, { "epoch": 2.6227968506454578, "grad_norm": 0.13767750561237335, "learning_rate": 8.57118646280411e-06, "loss": 0.0243, "step": 131250 }, { "epoch": 2.6229966827864595, "grad_norm": 0.13546256721019745, "learning_rate": 8.565401482698027e-06, "loss": 0.0225, "step": 131260 }, { "epoch": 2.6231965149274608, "grad_norm": 0.19403769075870514, "learning_rate": 8.559618272601877e-06, "loss": 0.0196, "step": 131270 }, { "epoch": 2.6233963470684625, "grad_norm": 0.10647568106651306, "learning_rate": 8.553836832762707e-06, "loss": 0.0253, "step": 131280 }, { "epoch": 2.623596179209464, "grad_norm": 0.12827788293361664, "learning_rate": 8.548057163427504e-06, "loss": 0.0239, "step": 131290 }, { "epoch": 2.6237960113504655, "grad_norm": 0.11036910861730576, "learning_rate": 8.542279264843117e-06, "loss": 0.0178, "step": 131300 }, { "epoch": 2.623995843491467, "grad_norm": 0.193570077419281, "learning_rate": 8.536503137256419e-06, "loss": 0.0171, "step": 131310 }, { "epoch": 2.6241956756324685, "grad_norm": 0.1419793665409088, "learning_rate": 8.530728780914143e-06, "loss": 0.0208, "step": 131320 }, { "epoch": 2.62439550777347, "grad_norm": 0.12993668019771576, "learning_rate": 8.524956196062938e-06, "loss": 0.0251, "step": 131330 }, { "epoch": 2.624595339914472, "grad_norm": 0.09253353625535965, "learning_rate": 8.519185382949406e-06, "loss": 0.0314, "step": 131340 }, { "epoch": 2.6247951720554736, "grad_norm": 0.10259465873241425, "learning_rate": 8.513416341820096e-06, "loss": 0.0248, "step": 131350 }, { "epoch": 2.624995004196475, "grad_norm": 0.12873238325119019, "learning_rate": 8.507649072921409e-06, "loss": 0.0229, "step": 131360 }, { "epoch": 2.6251948363374766, "grad_norm": 0.13693876564502716, "learning_rate": 8.501883576499735e-06, "loss": 0.0326, "step": 131370 }, { "epoch": 2.625394668478478, "grad_norm": 0.10720854252576828, "learning_rate": 8.49611985280136e-06, "loss": 0.022, "step": 131380 }, { "epoch": 2.6255945006194796, "grad_norm": 0.3432600796222687, "learning_rate": 8.490357902072492e-06, "loss": 0.0242, "step": 131390 }, { "epoch": 2.6257943327604814, "grad_norm": 0.07397737354040146, "learning_rate": 8.48459772455929e-06, "loss": 0.0234, "step": 131400 }, { "epoch": 2.6259941649014826, "grad_norm": 0.1819869577884674, "learning_rate": 8.478839320507798e-06, "loss": 0.0171, "step": 131410 }, { "epoch": 2.6261939970424844, "grad_norm": 0.11185863614082336, "learning_rate": 8.473082690164019e-06, "loss": 0.016, "step": 131420 }, { "epoch": 2.6263938291834856, "grad_norm": 0.15745697915554047, "learning_rate": 8.467327833773858e-06, "loss": 0.0225, "step": 131430 }, { "epoch": 2.6265936613244873, "grad_norm": 0.14480827748775482, "learning_rate": 8.461574751583152e-06, "loss": 0.0296, "step": 131440 }, { "epoch": 2.626793493465489, "grad_norm": 0.15529991686344147, "learning_rate": 8.455823443837669e-06, "loss": 0.0283, "step": 131450 }, { "epoch": 2.626993325606491, "grad_norm": 0.1114521399140358, "learning_rate": 8.450073910783097e-06, "loss": 0.0243, "step": 131460 }, { "epoch": 2.627193157747492, "grad_norm": 0.11010933667421341, "learning_rate": 8.444326152665016e-06, "loss": 0.0195, "step": 131470 }, { "epoch": 2.627392989888494, "grad_norm": 0.10494652390480042, "learning_rate": 8.438580169728993e-06, "loss": 0.0358, "step": 131480 }, { "epoch": 2.627592822029495, "grad_norm": 0.07911214977502823, "learning_rate": 8.432835962220486e-06, "loss": 0.0226, "step": 131490 }, { "epoch": 2.627792654170497, "grad_norm": 0.10107280313968658, "learning_rate": 8.427093530384856e-06, "loss": 0.0262, "step": 131500 }, { "epoch": 2.6279924863114985, "grad_norm": 0.1695302426815033, "learning_rate": 8.421352874467413e-06, "loss": 0.0243, "step": 131510 }, { "epoch": 2.6281923184525, "grad_norm": 0.16213053464889526, "learning_rate": 8.415613994713394e-06, "loss": 0.0245, "step": 131520 }, { "epoch": 2.6283921505935015, "grad_norm": 0.11955402791500092, "learning_rate": 8.409876891367958e-06, "loss": 0.0212, "step": 131530 }, { "epoch": 2.628591982734503, "grad_norm": 0.12103725969791412, "learning_rate": 8.404141564676171e-06, "loss": 0.0341, "step": 131540 }, { "epoch": 2.6287918148755045, "grad_norm": 0.10511592775583267, "learning_rate": 8.398408014883048e-06, "loss": 0.0176, "step": 131550 }, { "epoch": 2.6289916470165062, "grad_norm": 0.08575693517923355, "learning_rate": 8.392676242233504e-06, "loss": 0.0257, "step": 131560 }, { "epoch": 2.629191479157508, "grad_norm": 0.4056098461151123, "learning_rate": 8.386946246972398e-06, "loss": 0.0262, "step": 131570 }, { "epoch": 2.6293913112985092, "grad_norm": 0.11422059684991837, "learning_rate": 8.381218029344506e-06, "loss": 0.0218, "step": 131580 }, { "epoch": 2.629591143439511, "grad_norm": 0.13115079700946808, "learning_rate": 8.375491589594537e-06, "loss": 0.0283, "step": 131590 }, { "epoch": 2.6297909755805122, "grad_norm": 0.0774451419711113, "learning_rate": 8.369766927967082e-06, "loss": 0.0196, "step": 131600 }, { "epoch": 2.629990807721514, "grad_norm": 0.15823860466480255, "learning_rate": 8.364044044706697e-06, "loss": 0.0272, "step": 131610 }, { "epoch": 2.6301906398625157, "grad_norm": 0.16804702579975128, "learning_rate": 8.35832294005789e-06, "loss": 0.0227, "step": 131620 }, { "epoch": 2.630390472003517, "grad_norm": 0.1256500482559204, "learning_rate": 8.352603614265014e-06, "loss": 0.0253, "step": 131630 }, { "epoch": 2.6305903041445187, "grad_norm": 0.08178607374429703, "learning_rate": 8.346886067572402e-06, "loss": 0.0223, "step": 131640 }, { "epoch": 2.63079013628552, "grad_norm": 0.08153239637613297, "learning_rate": 8.3411703002243e-06, "loss": 0.0209, "step": 131650 }, { "epoch": 2.6309899684265217, "grad_norm": 0.13750672340393066, "learning_rate": 8.335456312464874e-06, "loss": 0.0255, "step": 131660 }, { "epoch": 2.6311898005675234, "grad_norm": 0.11937383562326431, "learning_rate": 8.329744104538212e-06, "loss": 0.0155, "step": 131670 }, { "epoch": 2.631389632708525, "grad_norm": 0.12434128671884537, "learning_rate": 8.32403367668833e-06, "loss": 0.0224, "step": 131680 }, { "epoch": 2.6315894648495264, "grad_norm": 0.1592472642660141, "learning_rate": 8.31832502915918e-06, "loss": 0.0228, "step": 131690 }, { "epoch": 2.631789296990528, "grad_norm": 0.12455374002456665, "learning_rate": 8.312618162194592e-06, "loss": 0.0237, "step": 131700 }, { "epoch": 2.6319891291315294, "grad_norm": 0.13963881134986877, "learning_rate": 8.30691307603838e-06, "loss": 0.0328, "step": 131710 }, { "epoch": 2.632188961272531, "grad_norm": 0.12934423983097076, "learning_rate": 8.301209770934271e-06, "loss": 0.0187, "step": 131720 }, { "epoch": 2.632388793413533, "grad_norm": 0.10871115326881409, "learning_rate": 8.295508247125854e-06, "loss": 0.0199, "step": 131730 }, { "epoch": 2.632588625554534, "grad_norm": 0.1203739270567894, "learning_rate": 8.28980850485671e-06, "loss": 0.0234, "step": 131740 }, { "epoch": 2.632788457695536, "grad_norm": 0.10762257874011993, "learning_rate": 8.284110544370338e-06, "loss": 0.0221, "step": 131750 }, { "epoch": 2.632988289836537, "grad_norm": 0.1349685937166214, "learning_rate": 8.27841436591012e-06, "loss": 0.0233, "step": 131760 }, { "epoch": 2.633188121977539, "grad_norm": 0.09434600174427032, "learning_rate": 8.2727199697194e-06, "loss": 0.0231, "step": 131770 }, { "epoch": 2.6333879541185405, "grad_norm": 0.14627155661582947, "learning_rate": 8.267027356041424e-06, "loss": 0.0188, "step": 131780 }, { "epoch": 2.633587786259542, "grad_norm": 0.13601678609848022, "learning_rate": 8.261336525119367e-06, "loss": 0.0237, "step": 131790 }, { "epoch": 2.6337876184005435, "grad_norm": 0.1227482482790947, "learning_rate": 8.255647477196343e-06, "loss": 0.0274, "step": 131800 }, { "epoch": 2.6339874505415453, "grad_norm": 0.17168429493904114, "learning_rate": 8.249960212515374e-06, "loss": 0.018, "step": 131810 }, { "epoch": 2.6341872826825465, "grad_norm": 0.10468766838312149, "learning_rate": 8.244274731319408e-06, "loss": 0.0207, "step": 131820 }, { "epoch": 2.6343871148235483, "grad_norm": 0.08914534002542496, "learning_rate": 8.238591033851312e-06, "loss": 0.0279, "step": 131830 }, { "epoch": 2.63458694696455, "grad_norm": 0.11304360628128052, "learning_rate": 8.232909120353893e-06, "loss": 0.0281, "step": 131840 }, { "epoch": 2.6347867791055513, "grad_norm": 0.10814603418111801, "learning_rate": 8.227228991069868e-06, "loss": 0.0333, "step": 131850 }, { "epoch": 2.634986611246553, "grad_norm": 0.12263013422489166, "learning_rate": 8.221550646241894e-06, "loss": 0.0194, "step": 131860 }, { "epoch": 2.6351864433875543, "grad_norm": 0.08058476448059082, "learning_rate": 8.215874086112508e-06, "loss": 0.0291, "step": 131870 }, { "epoch": 2.635386275528556, "grad_norm": 0.1368543952703476, "learning_rate": 8.210199310924216e-06, "loss": 0.0222, "step": 131880 }, { "epoch": 2.6355861076695577, "grad_norm": 0.08157849311828613, "learning_rate": 8.204526320919459e-06, "loss": 0.0207, "step": 131890 }, { "epoch": 2.635785939810559, "grad_norm": 0.10804662853479385, "learning_rate": 8.198855116340549e-06, "loss": 0.0164, "step": 131900 }, { "epoch": 2.6359857719515607, "grad_norm": 0.13321061432361603, "learning_rate": 8.193185697429756e-06, "loss": 0.0213, "step": 131910 }, { "epoch": 2.636185604092562, "grad_norm": 0.14025813341140747, "learning_rate": 8.187518064429266e-06, "loss": 0.0244, "step": 131920 }, { "epoch": 2.6363854362335637, "grad_norm": 0.2051362693309784, "learning_rate": 8.181852217581198e-06, "loss": 0.0236, "step": 131930 }, { "epoch": 2.6365852683745654, "grad_norm": 0.11896184086799622, "learning_rate": 8.176188157127573e-06, "loss": 0.0193, "step": 131940 }, { "epoch": 2.636785100515567, "grad_norm": 0.10809588432312012, "learning_rate": 8.170525883310365e-06, "loss": 0.0227, "step": 131950 }, { "epoch": 2.6369849326565684, "grad_norm": 0.12505070865154266, "learning_rate": 8.164865396371458e-06, "loss": 0.0238, "step": 131960 }, { "epoch": 2.63718476479757, "grad_norm": 0.09573009610176086, "learning_rate": 8.159206696552617e-06, "loss": 0.0273, "step": 131970 }, { "epoch": 2.6373845969385714, "grad_norm": 0.11978381127119064, "learning_rate": 8.15354978409562e-06, "loss": 0.0199, "step": 131980 }, { "epoch": 2.637584429079573, "grad_norm": 0.09854525327682495, "learning_rate": 8.147894659242111e-06, "loss": 0.0168, "step": 131990 }, { "epoch": 2.637784261220575, "grad_norm": 0.12097173184156418, "learning_rate": 8.142241322233651e-06, "loss": 0.0337, "step": 132000 }, { "epoch": 2.637984093361576, "grad_norm": 0.1477919965982437, "learning_rate": 8.136589773311725e-06, "loss": 0.0216, "step": 132010 }, { "epoch": 2.638183925502578, "grad_norm": 0.12590603530406952, "learning_rate": 8.130940012717813e-06, "loss": 0.0266, "step": 132020 }, { "epoch": 2.638383757643579, "grad_norm": 0.11276158690452576, "learning_rate": 8.12529204069321e-06, "loss": 0.0271, "step": 132030 }, { "epoch": 2.638583589784581, "grad_norm": 0.09828153997659683, "learning_rate": 8.119645857479202e-06, "loss": 0.0273, "step": 132040 }, { "epoch": 2.6387834219255826, "grad_norm": 0.11044483631849289, "learning_rate": 8.114001463316995e-06, "loss": 0.0209, "step": 132050 }, { "epoch": 2.6389832540665843, "grad_norm": 0.0994228720664978, "learning_rate": 8.10835885844769e-06, "loss": 0.0249, "step": 132060 }, { "epoch": 2.6391830862075856, "grad_norm": 0.10664104670286179, "learning_rate": 8.10271804311234e-06, "loss": 0.0159, "step": 132070 }, { "epoch": 2.6393829183485873, "grad_norm": 0.12551940977573395, "learning_rate": 8.097079017551907e-06, "loss": 0.025, "step": 132080 }, { "epoch": 2.6395827504895886, "grad_norm": 0.1304987370967865, "learning_rate": 8.091441782007297e-06, "loss": 0.0284, "step": 132090 }, { "epoch": 2.6397825826305903, "grad_norm": 0.11016152799129486, "learning_rate": 8.085806336719276e-06, "loss": 0.0239, "step": 132100 }, { "epoch": 2.639982414771592, "grad_norm": 0.10096252709627151, "learning_rate": 8.080172681928622e-06, "loss": 0.0232, "step": 132110 }, { "epoch": 2.6401822469125933, "grad_norm": 0.09496825933456421, "learning_rate": 8.074540817875998e-06, "loss": 0.0234, "step": 132120 }, { "epoch": 2.640382079053595, "grad_norm": 0.11054898798465729, "learning_rate": 8.068910744801955e-06, "loss": 0.0198, "step": 132130 }, { "epoch": 2.6405819111945963, "grad_norm": 0.0793592557311058, "learning_rate": 8.063282462947015e-06, "loss": 0.0186, "step": 132140 }, { "epoch": 2.640781743335598, "grad_norm": 0.1235157921910286, "learning_rate": 8.057655972551608e-06, "loss": 0.02, "step": 132150 }, { "epoch": 2.6409815754765997, "grad_norm": 0.11852218210697174, "learning_rate": 8.052031273856087e-06, "loss": 0.0284, "step": 132160 }, { "epoch": 2.6411814076176015, "grad_norm": 0.14272716641426086, "learning_rate": 8.046408367100734e-06, "loss": 0.0254, "step": 132170 }, { "epoch": 2.6413812397586027, "grad_norm": 0.10998518764972687, "learning_rate": 8.040787252525738e-06, "loss": 0.0194, "step": 132180 }, { "epoch": 2.6415810718996044, "grad_norm": 0.12595553696155548, "learning_rate": 8.035167930371228e-06, "loss": 0.0196, "step": 132190 }, { "epoch": 2.6417809040406057, "grad_norm": 0.14665238559246063, "learning_rate": 8.029550400877257e-06, "loss": 0.0313, "step": 132200 }, { "epoch": 2.6419807361816074, "grad_norm": 0.08471494913101196, "learning_rate": 8.023934664283789e-06, "loss": 0.0241, "step": 132210 }, { "epoch": 2.642180568322609, "grad_norm": 0.12061599642038345, "learning_rate": 8.018320720830719e-06, "loss": 0.0174, "step": 132220 }, { "epoch": 2.6423804004636104, "grad_norm": 0.11886730045080185, "learning_rate": 8.012708570757882e-06, "loss": 0.0288, "step": 132230 }, { "epoch": 2.642580232604612, "grad_norm": 0.11398506909608841, "learning_rate": 8.00709821430497e-06, "loss": 0.0176, "step": 132240 }, { "epoch": 2.6427800647456134, "grad_norm": 0.13898681104183197, "learning_rate": 8.001489651711696e-06, "loss": 0.0187, "step": 132250 }, { "epoch": 2.642979896886615, "grad_norm": 0.1348562091588974, "learning_rate": 7.995882883217642e-06, "loss": 0.0225, "step": 132260 }, { "epoch": 2.643179729027617, "grad_norm": 0.12771177291870117, "learning_rate": 7.990277909062294e-06, "loss": 0.0203, "step": 132270 }, { "epoch": 2.6433795611686186, "grad_norm": 0.14349541068077087, "learning_rate": 7.984674729485087e-06, "loss": 0.0261, "step": 132280 }, { "epoch": 2.64357939330962, "grad_norm": 0.1033916026353836, "learning_rate": 7.979073344725419e-06, "loss": 0.0254, "step": 132290 }, { "epoch": 2.6437792254506216, "grad_norm": 0.10478781163692474, "learning_rate": 7.973473755022527e-06, "loss": 0.0258, "step": 132300 }, { "epoch": 2.643979057591623, "grad_norm": 0.1385248899459839, "learning_rate": 7.967875960615629e-06, "loss": 0.0258, "step": 132310 }, { "epoch": 2.6441788897326246, "grad_norm": 0.17028900980949402, "learning_rate": 7.962279961743857e-06, "loss": 0.0315, "step": 132320 }, { "epoch": 2.6443787218736263, "grad_norm": 0.0867372676730156, "learning_rate": 7.956685758646259e-06, "loss": 0.0246, "step": 132330 }, { "epoch": 2.6445785540146276, "grad_norm": 0.10664060711860657, "learning_rate": 7.951093351561806e-06, "loss": 0.0245, "step": 132340 }, { "epoch": 2.6447783861556293, "grad_norm": 0.1096310243010521, "learning_rate": 7.945502740729405e-06, "loss": 0.0246, "step": 132350 }, { "epoch": 2.6449782182966306, "grad_norm": 0.1080409437417984, "learning_rate": 7.939913926387877e-06, "loss": 0.0225, "step": 132360 }, { "epoch": 2.6451780504376323, "grad_norm": 0.12711618840694427, "learning_rate": 7.93432690877594e-06, "loss": 0.0285, "step": 132370 }, { "epoch": 2.645377882578634, "grad_norm": 0.11887691915035248, "learning_rate": 7.92874168813229e-06, "loss": 0.0266, "step": 132380 }, { "epoch": 2.6455777147196358, "grad_norm": 0.10686219483613968, "learning_rate": 7.923158264695518e-06, "loss": 0.0195, "step": 132390 }, { "epoch": 2.645777546860637, "grad_norm": 0.17173007130622864, "learning_rate": 7.917576638704116e-06, "loss": 0.0236, "step": 132400 }, { "epoch": 2.6459773790016388, "grad_norm": 0.12396158277988434, "learning_rate": 7.911996810396538e-06, "loss": 0.0214, "step": 132410 }, { "epoch": 2.64617721114264, "grad_norm": 0.1277056783437729, "learning_rate": 7.906418780011132e-06, "loss": 0.0232, "step": 132420 }, { "epoch": 2.6463770432836418, "grad_norm": 0.10162249952554703, "learning_rate": 7.900842547786186e-06, "loss": 0.0252, "step": 132430 }, { "epoch": 2.6465768754246435, "grad_norm": 0.11926905810832977, "learning_rate": 7.895268113959914e-06, "loss": 0.0185, "step": 132440 }, { "epoch": 2.6467767075656448, "grad_norm": 0.08978797495365143, "learning_rate": 7.889695478770431e-06, "loss": 0.0185, "step": 132450 }, { "epoch": 2.6469765397066465, "grad_norm": 0.15512615442276, "learning_rate": 7.884124642455804e-06, "loss": 0.0216, "step": 132460 }, { "epoch": 2.6471763718476478, "grad_norm": 0.09932635724544525, "learning_rate": 7.878555605254e-06, "loss": 0.0191, "step": 132470 }, { "epoch": 2.6473762039886495, "grad_norm": 0.13178053498268127, "learning_rate": 7.872988367402922e-06, "loss": 0.0215, "step": 132480 }, { "epoch": 2.647576036129651, "grad_norm": 0.11615164577960968, "learning_rate": 7.867422929140405e-06, "loss": 0.018, "step": 132490 }, { "epoch": 2.647775868270653, "grad_norm": 0.16930387914180756, "learning_rate": 7.86185929070417e-06, "loss": 0.024, "step": 132500 }, { "epoch": 2.647975700411654, "grad_norm": 0.11082588881254196, "learning_rate": 7.856297452331879e-06, "loss": 0.0304, "step": 132510 }, { "epoch": 2.648175532552656, "grad_norm": 0.09307593107223511, "learning_rate": 7.850737414261167e-06, "loss": 0.0198, "step": 132520 }, { "epoch": 2.648375364693657, "grad_norm": 0.11266687512397766, "learning_rate": 7.845179176729511e-06, "loss": 0.0201, "step": 132530 }, { "epoch": 2.648575196834659, "grad_norm": 0.13026495277881622, "learning_rate": 7.839622739974356e-06, "loss": 0.0233, "step": 132540 }, { "epoch": 2.6487750289756606, "grad_norm": 0.08896251767873764, "learning_rate": 7.834068104233055e-06, "loss": 0.0223, "step": 132550 }, { "epoch": 2.648974861116662, "grad_norm": 0.14548882842063904, "learning_rate": 7.828515269742932e-06, "loss": 0.0212, "step": 132560 }, { "epoch": 2.6491746932576636, "grad_norm": 0.1022108644247055, "learning_rate": 7.822964236741148e-06, "loss": 0.0249, "step": 132570 }, { "epoch": 2.649374525398665, "grad_norm": 0.13100576400756836, "learning_rate": 7.817415005464852e-06, "loss": 0.0257, "step": 132580 }, { "epoch": 2.6495743575396666, "grad_norm": 0.11901865154504776, "learning_rate": 7.8118675761511e-06, "loss": 0.0269, "step": 132590 }, { "epoch": 2.6497741896806684, "grad_norm": 0.13480587303638458, "learning_rate": 7.806321949036855e-06, "loss": 0.0275, "step": 132600 }, { "epoch": 2.6499740218216696, "grad_norm": 0.12402968108654022, "learning_rate": 7.800778124359026e-06, "loss": 0.0207, "step": 132610 }, { "epoch": 2.6501738539626714, "grad_norm": 0.12815658748149872, "learning_rate": 7.795236102354436e-06, "loss": 0.0256, "step": 132620 }, { "epoch": 2.650373686103673, "grad_norm": 0.09079039841890335, "learning_rate": 7.78969588325984e-06, "loss": 0.0151, "step": 132630 }, { "epoch": 2.6505735182446744, "grad_norm": 0.1450161337852478, "learning_rate": 7.784157467311865e-06, "loss": 0.0226, "step": 132640 }, { "epoch": 2.650773350385676, "grad_norm": 0.10099650919437408, "learning_rate": 7.778620854747142e-06, "loss": 0.0193, "step": 132650 }, { "epoch": 2.650973182526678, "grad_norm": 0.11980137974023819, "learning_rate": 7.773086045802185e-06, "loss": 0.022, "step": 132660 }, { "epoch": 2.651173014667679, "grad_norm": 0.09836602956056595, "learning_rate": 7.76755304071341e-06, "loss": 0.0247, "step": 132670 }, { "epoch": 2.651372846808681, "grad_norm": 0.11805609613656998, "learning_rate": 7.762021839717171e-06, "loss": 0.0314, "step": 132680 }, { "epoch": 2.651572678949682, "grad_norm": 0.1194981187582016, "learning_rate": 7.75649244304979e-06, "loss": 0.0224, "step": 132690 }, { "epoch": 2.651772511090684, "grad_norm": 0.10794929414987564, "learning_rate": 7.75096485094744e-06, "loss": 0.021, "step": 132700 }, { "epoch": 2.6519723432316855, "grad_norm": 0.22414791584014893, "learning_rate": 7.745439063646254e-06, "loss": 0.0218, "step": 132710 }, { "epoch": 2.652172175372687, "grad_norm": 0.12405527383089066, "learning_rate": 7.739915081382288e-06, "loss": 0.0241, "step": 132720 }, { "epoch": 2.6523720075136885, "grad_norm": 0.12303472310304642, "learning_rate": 7.734392904391513e-06, "loss": 0.021, "step": 132730 }, { "epoch": 2.65257183965469, "grad_norm": 0.10183318704366684, "learning_rate": 7.728872532909832e-06, "loss": 0.0202, "step": 132740 }, { "epoch": 2.6527716717956915, "grad_norm": 0.14673982560634613, "learning_rate": 7.723353967173057e-06, "loss": 0.0268, "step": 132750 }, { "epoch": 2.6529715039366932, "grad_norm": 0.08350911736488342, "learning_rate": 7.71783720741695e-06, "loss": 0.0258, "step": 132760 }, { "epoch": 2.653171336077695, "grad_norm": 0.13303202390670776, "learning_rate": 7.71232225387714e-06, "loss": 0.0206, "step": 132770 }, { "epoch": 2.6533711682186962, "grad_norm": 0.1601274460554123, "learning_rate": 7.706809106789254e-06, "loss": 0.0261, "step": 132780 }, { "epoch": 2.653571000359698, "grad_norm": 0.09460406005382538, "learning_rate": 7.701297766388792e-06, "loss": 0.0262, "step": 132790 }, { "epoch": 2.6537708325006992, "grad_norm": 0.09689021110534668, "learning_rate": 7.695788232911178e-06, "loss": 0.0209, "step": 132800 }, { "epoch": 2.653970664641701, "grad_norm": 0.20242857933044434, "learning_rate": 7.690280506591775e-06, "loss": 0.035, "step": 132810 }, { "epoch": 2.6541704967827027, "grad_norm": 0.09858467429876328, "learning_rate": 7.68477458766586e-06, "loss": 0.0276, "step": 132820 }, { "epoch": 2.654370328923704, "grad_norm": 0.15284357964992523, "learning_rate": 7.679270476368638e-06, "loss": 0.0275, "step": 132830 }, { "epoch": 2.6545701610647057, "grad_norm": 0.13722127676010132, "learning_rate": 7.673768172935242e-06, "loss": 0.0192, "step": 132840 }, { "epoch": 2.654769993205707, "grad_norm": 0.10373887419700623, "learning_rate": 7.66826767760071e-06, "loss": 0.0294, "step": 132850 }, { "epoch": 2.6549698253467087, "grad_norm": 0.15717114508152008, "learning_rate": 7.662768990600017e-06, "loss": 0.0216, "step": 132860 }, { "epoch": 2.6551696574877104, "grad_norm": 0.1400129795074463, "learning_rate": 7.657272112168057e-06, "loss": 0.0243, "step": 132870 }, { "epoch": 2.655369489628712, "grad_norm": 0.14242562651634216, "learning_rate": 7.651777042539644e-06, "loss": 0.0232, "step": 132880 }, { "epoch": 2.6555693217697134, "grad_norm": 0.08949030935764313, "learning_rate": 7.646283781949538e-06, "loss": 0.0246, "step": 132890 }, { "epoch": 2.655769153910715, "grad_norm": 0.07395832985639572, "learning_rate": 7.640792330632363e-06, "loss": 0.0285, "step": 132900 }, { "epoch": 2.6559689860517164, "grad_norm": 0.13601815700531006, "learning_rate": 7.635302688822715e-06, "loss": 0.0241, "step": 132910 }, { "epoch": 2.656168818192718, "grad_norm": 0.0996197834610939, "learning_rate": 7.629814856755135e-06, "loss": 0.0285, "step": 132920 }, { "epoch": 2.65636865033372, "grad_norm": 0.10750597715377808, "learning_rate": 7.624328834664013e-06, "loss": 0.0222, "step": 132930 }, { "epoch": 2.656568482474721, "grad_norm": 0.12363208085298538, "learning_rate": 7.6188446227837165e-06, "loss": 0.0248, "step": 132940 }, { "epoch": 2.656768314615723, "grad_norm": 0.14342352747917175, "learning_rate": 7.613362221348513e-06, "loss": 0.0289, "step": 132950 }, { "epoch": 2.656968146756724, "grad_norm": 0.14100413024425507, "learning_rate": 7.607881630592628e-06, "loss": 0.0354, "step": 132960 }, { "epoch": 2.657167978897726, "grad_norm": 0.11568791419267654, "learning_rate": 7.602402850750151e-06, "loss": 0.0289, "step": 132970 }, { "epoch": 2.6573678110387275, "grad_norm": 0.12576326727867126, "learning_rate": 7.596925882055139e-06, "loss": 0.028, "step": 132980 }, { "epoch": 2.6575676431797293, "grad_norm": 0.08642605692148209, "learning_rate": 7.591450724741561e-06, "loss": 0.0253, "step": 132990 }, { "epoch": 2.6577674753207305, "grad_norm": 0.1352451741695404, "learning_rate": 7.585977379043296e-06, "loss": 0.023, "step": 133000 }, { "epoch": 2.6579673074617323, "grad_norm": 0.10064971446990967, "learning_rate": 7.580505845194158e-06, "loss": 0.0281, "step": 133010 }, { "epoch": 2.6581671396027335, "grad_norm": 0.15493318438529968, "learning_rate": 7.5750361234278865e-06, "loss": 0.0216, "step": 133020 }, { "epoch": 2.6583669717437353, "grad_norm": 0.08677944540977478, "learning_rate": 7.569568213978145e-06, "loss": 0.0274, "step": 133030 }, { "epoch": 2.658566803884737, "grad_norm": 0.1186484694480896, "learning_rate": 7.564102117078481e-06, "loss": 0.0191, "step": 133040 }, { "epoch": 2.6587666360257383, "grad_norm": 0.11752315610647202, "learning_rate": 7.5586378329624284e-06, "loss": 0.028, "step": 133050 }, { "epoch": 2.65896646816674, "grad_norm": 0.1061771884560585, "learning_rate": 7.553175361863413e-06, "loss": 0.021, "step": 133060 }, { "epoch": 2.6591663003077413, "grad_norm": 0.10992095619440079, "learning_rate": 7.547714704014758e-06, "loss": 0.0271, "step": 133070 }, { "epoch": 2.659366132448743, "grad_norm": 0.13009628653526306, "learning_rate": 7.5422558596497385e-06, "loss": 0.0227, "step": 133080 }, { "epoch": 2.6595659645897447, "grad_norm": 0.1169167011976242, "learning_rate": 7.536798829001557e-06, "loss": 0.0139, "step": 133090 }, { "epoch": 2.6597657967307464, "grad_norm": 0.2397877424955368, "learning_rate": 7.531343612303315e-06, "loss": 0.0184, "step": 133100 }, { "epoch": 2.6599656288717477, "grad_norm": 0.13481618463993073, "learning_rate": 7.5258902097880614e-06, "loss": 0.0237, "step": 133110 }, { "epoch": 2.6601654610127494, "grad_norm": 0.09197285771369934, "learning_rate": 7.520438621688752e-06, "loss": 0.0279, "step": 133120 }, { "epoch": 2.6603652931537507, "grad_norm": 0.11043555289506912, "learning_rate": 7.514988848238264e-06, "loss": 0.0288, "step": 133130 }, { "epoch": 2.6605651252947524, "grad_norm": 0.1593008041381836, "learning_rate": 7.509540889669403e-06, "loss": 0.018, "step": 133140 }, { "epoch": 2.660764957435754, "grad_norm": 0.16908057034015656, "learning_rate": 7.504094746214902e-06, "loss": 0.0298, "step": 133150 }, { "epoch": 2.6609647895767554, "grad_norm": 0.0932430699467659, "learning_rate": 7.498650418107417e-06, "loss": 0.0249, "step": 133160 }, { "epoch": 2.661164621717757, "grad_norm": 0.1408303827047348, "learning_rate": 7.493207905579497e-06, "loss": 0.0277, "step": 133170 }, { "epoch": 2.6613644538587584, "grad_norm": 0.13096703588962555, "learning_rate": 7.487767208863633e-06, "loss": 0.0227, "step": 133180 }, { "epoch": 2.66156428599976, "grad_norm": 0.11981142312288284, "learning_rate": 7.482328328192284e-06, "loss": 0.0199, "step": 133190 }, { "epoch": 2.661764118140762, "grad_norm": 0.1293202042579651, "learning_rate": 7.476891263797747e-06, "loss": 0.0262, "step": 133200 }, { "epoch": 2.6619639502817636, "grad_norm": 0.10647821426391602, "learning_rate": 7.471456015912298e-06, "loss": 0.0333, "step": 133210 }, { "epoch": 2.662163782422765, "grad_norm": 0.11429843306541443, "learning_rate": 7.466022584768123e-06, "loss": 0.0201, "step": 133220 }, { "epoch": 2.6623636145637666, "grad_norm": 0.09732088446617126, "learning_rate": 7.460590970597325e-06, "loss": 0.0259, "step": 133230 }, { "epoch": 2.662563446704768, "grad_norm": 0.13714760541915894, "learning_rate": 7.4551611736319305e-06, "loss": 0.0282, "step": 133240 }, { "epoch": 2.6627632788457696, "grad_norm": 0.1253490149974823, "learning_rate": 7.449733194103897e-06, "loss": 0.0248, "step": 133250 }, { "epoch": 2.6629631109867713, "grad_norm": 0.17878352105617523, "learning_rate": 7.444307032245102e-06, "loss": 0.0295, "step": 133260 }, { "epoch": 2.6631629431277726, "grad_norm": 0.10561936348676682, "learning_rate": 7.438882688287313e-06, "loss": 0.0248, "step": 133270 }, { "epoch": 2.6633627752687743, "grad_norm": 0.12372621893882751, "learning_rate": 7.43346016246228e-06, "loss": 0.0206, "step": 133280 }, { "epoch": 2.6635626074097756, "grad_norm": 0.11084972321987152, "learning_rate": 7.428039455001639e-06, "loss": 0.0281, "step": 133290 }, { "epoch": 2.6637624395507773, "grad_norm": 0.15478627383708954, "learning_rate": 7.42262056613694e-06, "loss": 0.0273, "step": 133300 }, { "epoch": 2.663962271691779, "grad_norm": 0.1031024158000946, "learning_rate": 7.417203496099656e-06, "loss": 0.0217, "step": 133310 }, { "epoch": 2.6641621038327807, "grad_norm": 0.12284477055072784, "learning_rate": 7.411788245121237e-06, "loss": 0.0186, "step": 133320 }, { "epoch": 2.664361935973782, "grad_norm": 0.11990875750780106, "learning_rate": 7.406374813432976e-06, "loss": 0.0253, "step": 133330 }, { "epoch": 2.6645617681147837, "grad_norm": 0.11481953412294388, "learning_rate": 7.400963201266137e-06, "loss": 0.0252, "step": 133340 }, { "epoch": 2.664761600255785, "grad_norm": 0.18403275310993195, "learning_rate": 7.395553408851891e-06, "loss": 0.023, "step": 133350 }, { "epoch": 2.6649614323967867, "grad_norm": 0.14479182660579681, "learning_rate": 7.390145436421336e-06, "loss": 0.0201, "step": 133360 }, { "epoch": 2.6651612645377885, "grad_norm": 0.10008738934993744, "learning_rate": 7.384739284205494e-06, "loss": 0.0265, "step": 133370 }, { "epoch": 2.6653610966787897, "grad_norm": 0.139683797955513, "learning_rate": 7.379334952435302e-06, "loss": 0.0318, "step": 133380 }, { "epoch": 2.6655609288197915, "grad_norm": 0.16518448293209076, "learning_rate": 7.373932441341619e-06, "loss": 0.0337, "step": 133390 }, { "epoch": 2.6657607609607927, "grad_norm": 0.29365256428718567, "learning_rate": 7.3685317511552445e-06, "loss": 0.0262, "step": 133400 }, { "epoch": 2.6659605931017945, "grad_norm": 0.09522748738527298, "learning_rate": 7.363132882106871e-06, "loss": 0.0206, "step": 133410 }, { "epoch": 2.666160425242796, "grad_norm": 0.12949879467487335, "learning_rate": 7.357735834427138e-06, "loss": 0.0189, "step": 133420 }, { "epoch": 2.6663602573837974, "grad_norm": 0.12249269336462021, "learning_rate": 7.352340608346603e-06, "loss": 0.0262, "step": 133430 }, { "epoch": 2.666560089524799, "grad_norm": 0.14961856603622437, "learning_rate": 7.346947204095717e-06, "loss": 0.0245, "step": 133440 }, { "epoch": 2.666759921665801, "grad_norm": 0.09939635545015335, "learning_rate": 7.341555621904883e-06, "loss": 0.0208, "step": 133450 }, { "epoch": 2.666959753806802, "grad_norm": 0.09775736927986145, "learning_rate": 7.336165862004446e-06, "loss": 0.0197, "step": 133460 }, { "epoch": 2.667159585947804, "grad_norm": 0.0859234407544136, "learning_rate": 7.330777924624621e-06, "loss": 0.0204, "step": 133470 }, { "epoch": 2.6673594180888056, "grad_norm": 0.07490785419940948, "learning_rate": 7.325391809995574e-06, "loss": 0.0269, "step": 133480 }, { "epoch": 2.667559250229807, "grad_norm": 0.10092869400978088, "learning_rate": 7.320007518347393e-06, "loss": 0.0239, "step": 133490 }, { "epoch": 2.6677590823708086, "grad_norm": 0.1300712376832962, "learning_rate": 7.314625049910084e-06, "loss": 0.028, "step": 133500 }, { "epoch": 2.66795891451181, "grad_norm": 0.13540171086788177, "learning_rate": 7.309244404913579e-06, "loss": 0.0203, "step": 133510 }, { "epoch": 2.6681587466528116, "grad_norm": 0.1178460568189621, "learning_rate": 7.303865583587726e-06, "loss": 0.0196, "step": 133520 }, { "epoch": 2.6683585787938133, "grad_norm": 0.11665946245193481, "learning_rate": 7.2984885861623165e-06, "loss": 0.0214, "step": 133530 }, { "epoch": 2.6685584109348146, "grad_norm": 0.1285640001296997, "learning_rate": 7.293113412867003e-06, "loss": 0.0261, "step": 133540 }, { "epoch": 2.6687582430758163, "grad_norm": 0.0952509269118309, "learning_rate": 7.287740063931442e-06, "loss": 0.0179, "step": 133550 }, { "epoch": 2.6689580752168176, "grad_norm": 0.0949121043086052, "learning_rate": 7.282368539585171e-06, "loss": 0.019, "step": 133560 }, { "epoch": 2.6691579073578193, "grad_norm": 0.07772237062454224, "learning_rate": 7.27699884005763e-06, "loss": 0.0253, "step": 133570 }, { "epoch": 2.669357739498821, "grad_norm": 0.11701009422540665, "learning_rate": 7.271630965578202e-06, "loss": 0.0192, "step": 133580 }, { "epoch": 2.6695575716398228, "grad_norm": 0.09116626530885696, "learning_rate": 7.266264916376225e-06, "loss": 0.0218, "step": 133590 }, { "epoch": 2.669757403780824, "grad_norm": 0.1218109056353569, "learning_rate": 7.2609006926809e-06, "loss": 0.0262, "step": 133600 }, { "epoch": 2.6699572359218258, "grad_norm": 0.11447872221469879, "learning_rate": 7.255538294721381e-06, "loss": 0.0333, "step": 133610 }, { "epoch": 2.670157068062827, "grad_norm": 0.13392393290996552, "learning_rate": 7.250177722726742e-06, "loss": 0.0278, "step": 133620 }, { "epoch": 2.6703569002038288, "grad_norm": 0.14103808999061584, "learning_rate": 7.244818976925976e-06, "loss": 0.025, "step": 133630 }, { "epoch": 2.6705567323448305, "grad_norm": 0.14717483520507812, "learning_rate": 7.239462057547996e-06, "loss": 0.0226, "step": 133640 }, { "epoch": 2.6707565644858318, "grad_norm": 0.10638108104467392, "learning_rate": 7.234106964821652e-06, "loss": 0.0222, "step": 133650 }, { "epoch": 2.6709563966268335, "grad_norm": 0.1237688735127449, "learning_rate": 7.228753698975699e-06, "loss": 0.0226, "step": 133660 }, { "epoch": 2.6711562287678348, "grad_norm": 0.136200413107872, "learning_rate": 7.223402260238793e-06, "loss": 0.0232, "step": 133670 }, { "epoch": 2.6713560609088365, "grad_norm": 0.15633457899093628, "learning_rate": 7.218052648839574e-06, "loss": 0.029, "step": 133680 }, { "epoch": 2.671555893049838, "grad_norm": 0.10893792659044266, "learning_rate": 7.212704865006564e-06, "loss": 0.0205, "step": 133690 }, { "epoch": 2.67175572519084, "grad_norm": 0.12015233933925629, "learning_rate": 7.207358908968187e-06, "loss": 0.0249, "step": 133700 }, { "epoch": 2.671955557331841, "grad_norm": 0.10608579218387604, "learning_rate": 7.20201478095282e-06, "loss": 0.0232, "step": 133710 }, { "epoch": 2.672155389472843, "grad_norm": 0.14569805562496185, "learning_rate": 7.196672481188765e-06, "loss": 0.0242, "step": 133720 }, { "epoch": 2.672355221613844, "grad_norm": 0.1333387792110443, "learning_rate": 7.1913320099042285e-06, "loss": 0.0268, "step": 133730 }, { "epoch": 2.672555053754846, "grad_norm": 0.08622033149003983, "learning_rate": 7.185993367327348e-06, "loss": 0.0297, "step": 133740 }, { "epoch": 2.6727548858958476, "grad_norm": 0.1246996819972992, "learning_rate": 7.180656553686177e-06, "loss": 0.0259, "step": 133750 }, { "epoch": 2.672954718036849, "grad_norm": 0.12554802000522614, "learning_rate": 7.175321569208693e-06, "loss": 0.0269, "step": 133760 }, { "epoch": 2.6731545501778506, "grad_norm": 0.1702062487602234, "learning_rate": 7.1699884141228076e-06, "loss": 0.0305, "step": 133770 }, { "epoch": 2.673354382318852, "grad_norm": 0.08798014372587204, "learning_rate": 7.164657088656329e-06, "loss": 0.0219, "step": 133780 }, { "epoch": 2.6735542144598536, "grad_norm": 0.2333514243364334, "learning_rate": 7.159327593037013e-06, "loss": 0.0327, "step": 133790 }, { "epoch": 2.6737540466008554, "grad_norm": 0.1348133534193039, "learning_rate": 7.1539999274925165e-06, "loss": 0.0214, "step": 133800 }, { "epoch": 2.673953878741857, "grad_norm": 0.10511668771505356, "learning_rate": 7.14867409225044e-06, "loss": 0.0322, "step": 133810 }, { "epoch": 2.6741537108828584, "grad_norm": 0.08144327998161316, "learning_rate": 7.143350087538281e-06, "loss": 0.02, "step": 133820 }, { "epoch": 2.67435354302386, "grad_norm": 0.29195284843444824, "learning_rate": 7.138027913583484e-06, "loss": 0.0213, "step": 133830 }, { "epoch": 2.6745533751648614, "grad_norm": 0.1279793530702591, "learning_rate": 7.1327075706133896e-06, "loss": 0.024, "step": 133840 }, { "epoch": 2.674753207305863, "grad_norm": 0.15103837847709656, "learning_rate": 7.127389058855266e-06, "loss": 0.0253, "step": 133850 }, { "epoch": 2.674953039446865, "grad_norm": 0.12299846857786179, "learning_rate": 7.122072378536343e-06, "loss": 0.0273, "step": 133860 }, { "epoch": 2.675152871587866, "grad_norm": 0.13049378991127014, "learning_rate": 7.116757529883705e-06, "loss": 0.0275, "step": 133870 }, { "epoch": 2.675352703728868, "grad_norm": 0.13047553598880768, "learning_rate": 7.11144451312441e-06, "loss": 0.0192, "step": 133880 }, { "epoch": 2.675552535869869, "grad_norm": 0.13790345191955566, "learning_rate": 7.1061333284854145e-06, "loss": 0.031, "step": 133890 }, { "epoch": 2.675752368010871, "grad_norm": 0.10839887708425522, "learning_rate": 7.100823976193604e-06, "loss": 0.0264, "step": 133900 }, { "epoch": 2.6759522001518725, "grad_norm": 0.12768585979938507, "learning_rate": 7.095516456475787e-06, "loss": 0.0249, "step": 133910 }, { "epoch": 2.6761520322928742, "grad_norm": 0.1325577199459076, "learning_rate": 7.090210769558692e-06, "loss": 0.0247, "step": 133920 }, { "epoch": 2.6763518644338755, "grad_norm": 0.15969817340373993, "learning_rate": 7.084906915668971e-06, "loss": 0.0223, "step": 133930 }, { "epoch": 2.6765516965748772, "grad_norm": 0.12415461987257004, "learning_rate": 7.07960489503317e-06, "loss": 0.0252, "step": 133940 }, { "epoch": 2.6767515287158785, "grad_norm": 0.10581902414560318, "learning_rate": 7.07430470787781e-06, "loss": 0.0185, "step": 133950 }, { "epoch": 2.6769513608568802, "grad_norm": 0.13595910370349884, "learning_rate": 7.069006354429308e-06, "loss": 0.0333, "step": 133960 }, { "epoch": 2.677151192997882, "grad_norm": 0.12836837768554688, "learning_rate": 7.063709834913973e-06, "loss": 0.0247, "step": 133970 }, { "epoch": 2.6773510251388832, "grad_norm": 0.17416171729564667, "learning_rate": 7.0584151495580734e-06, "loss": 0.0259, "step": 133980 }, { "epoch": 2.677550857279885, "grad_norm": 0.1914055496454239, "learning_rate": 7.053122298587811e-06, "loss": 0.0227, "step": 133990 }, { "epoch": 2.6777506894208862, "grad_norm": 0.08919639885425568, "learning_rate": 7.047831282229256e-06, "loss": 0.0248, "step": 134000 }, { "epoch": 2.677950521561888, "grad_norm": 0.10044050216674805, "learning_rate": 7.042542100708444e-06, "loss": 0.0272, "step": 134010 }, { "epoch": 2.6781503537028897, "grad_norm": 0.13103224337100983, "learning_rate": 7.037254754251316e-06, "loss": 0.0242, "step": 134020 }, { "epoch": 2.6783501858438914, "grad_norm": 0.17995993793010712, "learning_rate": 7.031969243083741e-06, "loss": 0.0256, "step": 134030 }, { "epoch": 2.6785500179848927, "grad_norm": 0.09295245260000229, "learning_rate": 7.026685567431501e-06, "loss": 0.0293, "step": 134040 }, { "epoch": 2.6787498501258944, "grad_norm": 0.13375453650951385, "learning_rate": 7.0214037275203144e-06, "loss": 0.0187, "step": 134050 }, { "epoch": 2.6789496822668957, "grad_norm": 0.14748899638652802, "learning_rate": 7.016123723575818e-06, "loss": 0.024, "step": 134060 }, { "epoch": 2.6791495144078974, "grad_norm": 0.10699934512376785, "learning_rate": 7.010845555823531e-06, "loss": 0.0255, "step": 134070 }, { "epoch": 2.679349346548899, "grad_norm": 0.150991290807724, "learning_rate": 7.005569224488956e-06, "loss": 0.0275, "step": 134080 }, { "epoch": 2.6795491786899004, "grad_norm": 0.08870486170053482, "learning_rate": 7.00029472979749e-06, "loss": 0.0252, "step": 134090 }, { "epoch": 2.679749010830902, "grad_norm": 0.10494372248649597, "learning_rate": 6.995022071974433e-06, "loss": 0.0242, "step": 134100 }, { "epoch": 2.6799488429719034, "grad_norm": 0.14583179354667664, "learning_rate": 6.98975125124503e-06, "loss": 0.0377, "step": 134110 }, { "epoch": 2.680148675112905, "grad_norm": 0.13563266396522522, "learning_rate": 6.984482267834441e-06, "loss": 0.0236, "step": 134120 }, { "epoch": 2.680348507253907, "grad_norm": 0.1044396385550499, "learning_rate": 6.979215121967747e-06, "loss": 0.0238, "step": 134130 }, { "epoch": 2.6805483393949086, "grad_norm": 0.10786129534244537, "learning_rate": 6.973949813869951e-06, "loss": 0.0235, "step": 134140 }, { "epoch": 2.68074817153591, "grad_norm": 0.10985566675662994, "learning_rate": 6.968686343765974e-06, "loss": 0.0354, "step": 134150 }, { "epoch": 2.6809480036769116, "grad_norm": 0.103366918861866, "learning_rate": 6.9634247118806674e-06, "loss": 0.0166, "step": 134160 }, { "epoch": 2.681147835817913, "grad_norm": 0.10240162163972855, "learning_rate": 6.9581649184387975e-06, "loss": 0.0205, "step": 134170 }, { "epoch": 2.6813476679589145, "grad_norm": 0.11217319220304489, "learning_rate": 6.952906963665057e-06, "loss": 0.0246, "step": 134180 }, { "epoch": 2.6815475000999163, "grad_norm": 0.1256357878446579, "learning_rate": 6.947650847784043e-06, "loss": 0.0313, "step": 134190 }, { "epoch": 2.6817473322409175, "grad_norm": 0.09851554781198502, "learning_rate": 6.9423965710203095e-06, "loss": 0.0272, "step": 134200 }, { "epoch": 2.6819471643819193, "grad_norm": 0.17918206751346588, "learning_rate": 6.937144133598272e-06, "loss": 0.0205, "step": 134210 }, { "epoch": 2.6821469965229205, "grad_norm": 0.11394688487052917, "learning_rate": 6.93189353574234e-06, "loss": 0.0247, "step": 134220 }, { "epoch": 2.6823468286639223, "grad_norm": 0.13811209797859192, "learning_rate": 6.926644777676811e-06, "loss": 0.0155, "step": 134230 }, { "epoch": 2.682546660804924, "grad_norm": 0.11651995033025742, "learning_rate": 6.921397859625872e-06, "loss": 0.0353, "step": 134240 }, { "epoch": 2.6827464929459253, "grad_norm": 0.09927159547805786, "learning_rate": 6.916152781813673e-06, "loss": 0.0244, "step": 134250 }, { "epoch": 2.682946325086927, "grad_norm": 0.10362054407596588, "learning_rate": 6.9109095444643e-06, "loss": 0.023, "step": 134260 }, { "epoch": 2.6831461572279283, "grad_norm": 0.1299118548631668, "learning_rate": 6.905668147801703e-06, "loss": 0.0181, "step": 134270 }, { "epoch": 2.68334598936893, "grad_norm": 0.13057540357112885, "learning_rate": 6.900428592049796e-06, "loss": 0.0233, "step": 134280 }, { "epoch": 2.6835458215099317, "grad_norm": 0.11596807092428207, "learning_rate": 6.8951908774324005e-06, "loss": 0.0243, "step": 134290 }, { "epoch": 2.6837456536509334, "grad_norm": 0.11780812591314316, "learning_rate": 6.889955004173266e-06, "loss": 0.0192, "step": 134300 }, { "epoch": 2.6839454857919347, "grad_norm": 0.12151793390512466, "learning_rate": 6.884720972496061e-06, "loss": 0.0211, "step": 134310 }, { "epoch": 2.6841453179329364, "grad_norm": 0.10777503997087479, "learning_rate": 6.879488782624366e-06, "loss": 0.0362, "step": 134320 }, { "epoch": 2.6843451500739377, "grad_norm": 0.11145920306444168, "learning_rate": 6.874258434781711e-06, "loss": 0.02, "step": 134330 }, { "epoch": 2.6845449822149394, "grad_norm": 0.13186106085777283, "learning_rate": 6.869029929191484e-06, "loss": 0.0251, "step": 134340 }, { "epoch": 2.684744814355941, "grad_norm": 0.10613809525966644, "learning_rate": 6.863803266077079e-06, "loss": 0.0246, "step": 134350 }, { "epoch": 2.6849446464969424, "grad_norm": 0.142470583319664, "learning_rate": 6.858578445661768e-06, "loss": 0.0251, "step": 134360 }, { "epoch": 2.685144478637944, "grad_norm": 0.11219625920057297, "learning_rate": 6.853355468168715e-06, "loss": 0.0203, "step": 134370 }, { "epoch": 2.6853443107789454, "grad_norm": 0.10794913023710251, "learning_rate": 6.84813433382106e-06, "loss": 0.0174, "step": 134380 }, { "epoch": 2.685544142919947, "grad_norm": 0.10084924846887589, "learning_rate": 6.84291504284183e-06, "loss": 0.0251, "step": 134390 }, { "epoch": 2.685743975060949, "grad_norm": 0.18527847528457642, "learning_rate": 6.837697595453985e-06, "loss": 0.023, "step": 134400 }, { "epoch": 2.6859438072019506, "grad_norm": 0.1210751011967659, "learning_rate": 6.832481991880413e-06, "loss": 0.0205, "step": 134410 }, { "epoch": 2.686143639342952, "grad_norm": 0.11920776218175888, "learning_rate": 6.827268232343903e-06, "loss": 0.02, "step": 134420 }, { "epoch": 2.6863434714839536, "grad_norm": 0.18271777033805847, "learning_rate": 6.8220563170671885e-06, "loss": 0.0229, "step": 134430 }, { "epoch": 2.686543303624955, "grad_norm": 0.11246901750564575, "learning_rate": 6.816846246272906e-06, "loss": 0.0297, "step": 134440 }, { "epoch": 2.6867431357659566, "grad_norm": 0.11604563146829605, "learning_rate": 6.811638020183619e-06, "loss": 0.0244, "step": 134450 }, { "epoch": 2.6869429679069583, "grad_norm": 0.18006516993045807, "learning_rate": 6.806431639021832e-06, "loss": 0.0214, "step": 134460 }, { "epoch": 2.6871428000479596, "grad_norm": 0.1178349107503891, "learning_rate": 6.801227103009922e-06, "loss": 0.0218, "step": 134470 }, { "epoch": 2.6873426321889613, "grad_norm": 0.159502774477005, "learning_rate": 6.796024412370222e-06, "loss": 0.0232, "step": 134480 }, { "epoch": 2.6875424643299626, "grad_norm": 0.18285587430000305, "learning_rate": 6.790823567325011e-06, "loss": 0.0254, "step": 134490 }, { "epoch": 2.6877422964709643, "grad_norm": 0.13589109480381012, "learning_rate": 6.785624568096432e-06, "loss": 0.0253, "step": 134500 }, { "epoch": 2.687942128611966, "grad_norm": 0.16975900530815125, "learning_rate": 6.780427414906582e-06, "loss": 0.0222, "step": 134510 }, { "epoch": 2.6881419607529677, "grad_norm": 0.13908465206623077, "learning_rate": 6.775232107977475e-06, "loss": 0.0331, "step": 134520 }, { "epoch": 2.688341792893969, "grad_norm": 0.1311616450548172, "learning_rate": 6.770038647531046e-06, "loss": 0.0221, "step": 134530 }, { "epoch": 2.6885416250349707, "grad_norm": 0.1283283829689026, "learning_rate": 6.764847033789151e-06, "loss": 0.0193, "step": 134540 }, { "epoch": 2.688741457175972, "grad_norm": 0.09554539620876312, "learning_rate": 6.759657266973562e-06, "loss": 0.0289, "step": 134550 }, { "epoch": 2.6889412893169737, "grad_norm": 0.0895858034491539, "learning_rate": 6.754469347305998e-06, "loss": 0.0255, "step": 134560 }, { "epoch": 2.6891411214579755, "grad_norm": 0.10806594789028168, "learning_rate": 6.7492832750080285e-06, "loss": 0.0262, "step": 134570 }, { "epoch": 2.6893409535989767, "grad_norm": 0.14603455364704132, "learning_rate": 6.744099050301239e-06, "loss": 0.0266, "step": 134580 }, { "epoch": 2.6895407857399785, "grad_norm": 0.0646880567073822, "learning_rate": 6.738916673407075e-06, "loss": 0.0262, "step": 134590 }, { "epoch": 2.6897406178809797, "grad_norm": 0.12295863032341003, "learning_rate": 6.733736144546926e-06, "loss": 0.0293, "step": 134600 }, { "epoch": 2.6899404500219815, "grad_norm": 0.13383200764656067, "learning_rate": 6.728557463942065e-06, "loss": 0.0188, "step": 134610 }, { "epoch": 2.690140282162983, "grad_norm": 0.12817220389842987, "learning_rate": 6.7233806318137475e-06, "loss": 0.0184, "step": 134620 }, { "epoch": 2.690340114303985, "grad_norm": 0.12952512502670288, "learning_rate": 6.718205648383119e-06, "loss": 0.0256, "step": 134630 }, { "epoch": 2.690539946444986, "grad_norm": 0.15709014236927032, "learning_rate": 6.7130325138712255e-06, "loss": 0.0232, "step": 134640 }, { "epoch": 2.690739778585988, "grad_norm": 0.12004102021455765, "learning_rate": 6.7078612284990576e-06, "loss": 0.0262, "step": 134650 }, { "epoch": 2.690939610726989, "grad_norm": 0.10407353192567825, "learning_rate": 6.702691792487531e-06, "loss": 0.0215, "step": 134660 }, { "epoch": 2.691139442867991, "grad_norm": 0.09371040761470795, "learning_rate": 6.697524206057471e-06, "loss": 0.0199, "step": 134670 }, { "epoch": 2.6913392750089926, "grad_norm": 0.1811968982219696, "learning_rate": 6.6923584694296275e-06, "loss": 0.0297, "step": 134680 }, { "epoch": 2.691539107149994, "grad_norm": 0.08902882784605026, "learning_rate": 6.687194582824674e-06, "loss": 0.0189, "step": 134690 }, { "epoch": 2.6917389392909956, "grad_norm": 0.1487075537443161, "learning_rate": 6.682032546463196e-06, "loss": 0.0264, "step": 134700 }, { "epoch": 2.691938771431997, "grad_norm": 0.0874486118555069, "learning_rate": 6.676872360565711e-06, "loss": 0.0284, "step": 134710 }, { "epoch": 2.6921386035729986, "grad_norm": 0.1131414920091629, "learning_rate": 6.671714025352654e-06, "loss": 0.019, "step": 134720 }, { "epoch": 2.6923384357140003, "grad_norm": 0.0897703617811203, "learning_rate": 6.666557541044383e-06, "loss": 0.0275, "step": 134730 }, { "epoch": 2.692538267855002, "grad_norm": 0.13564635813236237, "learning_rate": 6.661402907861164e-06, "loss": 0.0157, "step": 134740 }, { "epoch": 2.6927380999960033, "grad_norm": 0.15269772708415985, "learning_rate": 6.656250126023178e-06, "loss": 0.0263, "step": 134750 }, { "epoch": 2.692937932137005, "grad_norm": 0.10164368897676468, "learning_rate": 6.6510991957505875e-06, "loss": 0.023, "step": 134760 }, { "epoch": 2.6931377642780063, "grad_norm": 0.09836814552545547, "learning_rate": 6.645950117263394e-06, "loss": 0.031, "step": 134770 }, { "epoch": 2.693337596419008, "grad_norm": 0.14567001163959503, "learning_rate": 6.640802890781573e-06, "loss": 0.0295, "step": 134780 }, { "epoch": 2.6935374285600098, "grad_norm": 0.09652921557426453, "learning_rate": 6.6356575165249905e-06, "loss": 0.0239, "step": 134790 }, { "epoch": 2.693737260701011, "grad_norm": 0.12159938365221024, "learning_rate": 6.630513994713461e-06, "loss": 0.0239, "step": 134800 }, { "epoch": 2.6939370928420128, "grad_norm": 0.12949587404727936, "learning_rate": 6.625372325566698e-06, "loss": 0.0288, "step": 134810 }, { "epoch": 2.694136924983014, "grad_norm": 0.1119404137134552, "learning_rate": 6.620232509304347e-06, "loss": 0.0208, "step": 134820 }, { "epoch": 2.6943367571240158, "grad_norm": 0.11583144217729568, "learning_rate": 6.615094546145989e-06, "loss": 0.0208, "step": 134830 }, { "epoch": 2.6945365892650175, "grad_norm": 0.13256017863750458, "learning_rate": 6.609958436311064e-06, "loss": 0.0241, "step": 134840 }, { "epoch": 2.694736421406019, "grad_norm": 0.09105489403009415, "learning_rate": 6.60482418001902e-06, "loss": 0.0256, "step": 134850 }, { "epoch": 2.6949362535470205, "grad_norm": 0.13390403985977173, "learning_rate": 6.599691777489176e-06, "loss": 0.0226, "step": 134860 }, { "epoch": 2.695136085688022, "grad_norm": 0.11434328556060791, "learning_rate": 6.5945612289407545e-06, "loss": 0.0234, "step": 134870 }, { "epoch": 2.6953359178290235, "grad_norm": 0.13421379029750824, "learning_rate": 6.589432534592932e-06, "loss": 0.0294, "step": 134880 }, { "epoch": 2.695535749970025, "grad_norm": 0.30735817551612854, "learning_rate": 6.584305694664827e-06, "loss": 0.0214, "step": 134890 }, { "epoch": 2.695735582111027, "grad_norm": 0.0856892466545105, "learning_rate": 6.579180709375415e-06, "loss": 0.0205, "step": 134900 }, { "epoch": 2.695935414252028, "grad_norm": 0.11066148430109024, "learning_rate": 6.574057578943637e-06, "loss": 0.0188, "step": 134910 }, { "epoch": 2.69613524639303, "grad_norm": 0.13557814061641693, "learning_rate": 6.5689363035883414e-06, "loss": 0.0291, "step": 134920 }, { "epoch": 2.696335078534031, "grad_norm": 0.13784778118133545, "learning_rate": 6.563816883528301e-06, "loss": 0.0199, "step": 134930 }, { "epoch": 2.696534910675033, "grad_norm": 0.13246577978134155, "learning_rate": 6.558699318982209e-06, "loss": 0.0228, "step": 134940 }, { "epoch": 2.6967347428160346, "grad_norm": 0.12310956418514252, "learning_rate": 6.553583610168679e-06, "loss": 0.0255, "step": 134950 }, { "epoch": 2.696934574957036, "grad_norm": 0.1499951332807541, "learning_rate": 6.548469757306253e-06, "loss": 0.0239, "step": 134960 }, { "epoch": 2.6971344070980376, "grad_norm": 0.11487945914268494, "learning_rate": 6.543357760613355e-06, "loss": 0.0259, "step": 134970 }, { "epoch": 2.6973342392390394, "grad_norm": 0.09574203938245773, "learning_rate": 6.538247620308391e-06, "loss": 0.0216, "step": 134980 }, { "epoch": 2.6975340713800406, "grad_norm": 0.13622242212295532, "learning_rate": 6.533139336609651e-06, "loss": 0.022, "step": 134990 }, { "epoch": 2.6977339035210424, "grad_norm": 0.18320883810520172, "learning_rate": 6.528032909735354e-06, "loss": 0.0238, "step": 135000 }, { "epoch": 2.697933735662044, "grad_norm": 0.15588092803955078, "learning_rate": 6.522928339903606e-06, "loss": 0.0291, "step": 135010 }, { "epoch": 2.6981335678030454, "grad_norm": 0.11847775429487228, "learning_rate": 6.5178256273325074e-06, "loss": 0.0187, "step": 135020 }, { "epoch": 2.698333399944047, "grad_norm": 0.17124420404434204, "learning_rate": 6.512724772240025e-06, "loss": 0.0211, "step": 135030 }, { "epoch": 2.6985332320850484, "grad_norm": 0.13536658883094788, "learning_rate": 6.507625774844039e-06, "loss": 0.0306, "step": 135040 }, { "epoch": 2.69873306422605, "grad_norm": 0.09733138978481293, "learning_rate": 6.502528635362387e-06, "loss": 0.0219, "step": 135050 }, { "epoch": 2.698932896367052, "grad_norm": 0.11352783441543579, "learning_rate": 6.4974333540128e-06, "loss": 0.0216, "step": 135060 }, { "epoch": 2.699132728508053, "grad_norm": 0.09327352792024612, "learning_rate": 6.4923399310129475e-06, "loss": 0.0191, "step": 135070 }, { "epoch": 2.699332560649055, "grad_norm": 0.13270889222621918, "learning_rate": 6.487248366580406e-06, "loss": 0.0192, "step": 135080 }, { "epoch": 2.699532392790056, "grad_norm": 0.1438145786523819, "learning_rate": 6.482158660932675e-06, "loss": 0.0284, "step": 135090 }, { "epoch": 2.699732224931058, "grad_norm": 0.11519208550453186, "learning_rate": 6.4770708142871836e-06, "loss": 0.0317, "step": 135100 }, { "epoch": 2.6999320570720595, "grad_norm": 0.1377282738685608, "learning_rate": 6.471984826861277e-06, "loss": 0.0241, "step": 135110 }, { "epoch": 2.7001318892130612, "grad_norm": 0.1248648390173912, "learning_rate": 6.466900698872208e-06, "loss": 0.0264, "step": 135120 }, { "epoch": 2.7003317213540625, "grad_norm": 0.1358300745487213, "learning_rate": 6.461818430537181e-06, "loss": 0.0258, "step": 135130 }, { "epoch": 2.7005315534950642, "grad_norm": 0.11659244447946548, "learning_rate": 6.456738022073283e-06, "loss": 0.0249, "step": 135140 }, { "epoch": 2.7007313856360655, "grad_norm": 0.10136505216360092, "learning_rate": 6.45165947369753e-06, "loss": 0.0221, "step": 135150 }, { "epoch": 2.7009312177770672, "grad_norm": 0.10711664706468582, "learning_rate": 6.4465827856269034e-06, "loss": 0.0217, "step": 135160 }, { "epoch": 2.701131049918069, "grad_norm": 0.1099681556224823, "learning_rate": 6.441507958078241e-06, "loss": 0.0233, "step": 135170 }, { "epoch": 2.7013308820590702, "grad_norm": 0.06740471720695496, "learning_rate": 6.436434991268342e-06, "loss": 0.0301, "step": 135180 }, { "epoch": 2.701530714200072, "grad_norm": 0.14806291460990906, "learning_rate": 6.4313638854139045e-06, "loss": 0.021, "step": 135190 }, { "epoch": 2.7017305463410732, "grad_norm": 0.1593208909034729, "learning_rate": 6.426294640731567e-06, "loss": 0.0245, "step": 135200 }, { "epoch": 2.701930378482075, "grad_norm": 0.11215067654848099, "learning_rate": 6.4212272574378785e-06, "loss": 0.0219, "step": 135210 }, { "epoch": 2.7021302106230767, "grad_norm": 0.13664130866527557, "learning_rate": 6.416161735749304e-06, "loss": 0.0283, "step": 135220 }, { "epoch": 2.7023300427640784, "grad_norm": 0.1294083595275879, "learning_rate": 6.411098075882238e-06, "loss": 0.0169, "step": 135230 }, { "epoch": 2.7025298749050797, "grad_norm": 0.11983407288789749, "learning_rate": 6.406036278052973e-06, "loss": 0.025, "step": 135240 }, { "epoch": 2.7027297070460814, "grad_norm": 0.12806060910224915, "learning_rate": 6.400976342477761e-06, "loss": 0.0218, "step": 135250 }, { "epoch": 2.7029295391870827, "grad_norm": 0.09911922365427017, "learning_rate": 6.395918269372758e-06, "loss": 0.021, "step": 135260 }, { "epoch": 2.7031293713280844, "grad_norm": 0.1312742829322815, "learning_rate": 6.390862058954011e-06, "loss": 0.0221, "step": 135270 }, { "epoch": 2.703329203469086, "grad_norm": 0.20295387506484985, "learning_rate": 6.385807711437513e-06, "loss": 0.0242, "step": 135280 }, { "epoch": 2.7035290356100874, "grad_norm": 0.08940261602401733, "learning_rate": 6.3807552270392094e-06, "loss": 0.0265, "step": 135290 }, { "epoch": 2.703728867751089, "grad_norm": 0.11776300519704819, "learning_rate": 6.375704605974902e-06, "loss": 0.0227, "step": 135300 }, { "epoch": 2.7039286998920904, "grad_norm": 0.11249150335788727, "learning_rate": 6.370655848460355e-06, "loss": 0.026, "step": 135310 }, { "epoch": 2.704128532033092, "grad_norm": 0.13220980763435364, "learning_rate": 6.365608954711239e-06, "loss": 0.0201, "step": 135320 }, { "epoch": 2.704328364174094, "grad_norm": 0.07718874514102936, "learning_rate": 6.360563924943147e-06, "loss": 0.0211, "step": 135330 }, { "epoch": 2.7045281963150956, "grad_norm": 0.08711112290620804, "learning_rate": 6.355520759371597e-06, "loss": 0.026, "step": 135340 }, { "epoch": 2.704728028456097, "grad_norm": 0.09111825376749039, "learning_rate": 6.350479458212022e-06, "loss": 0.0245, "step": 135350 }, { "epoch": 2.7049278605970986, "grad_norm": 0.3759750425815582, "learning_rate": 6.3454400216797826e-06, "loss": 0.0242, "step": 135360 }, { "epoch": 2.7051276927381, "grad_norm": 0.12791723012924194, "learning_rate": 6.340402449990146e-06, "loss": 0.0235, "step": 135370 }, { "epoch": 2.7053275248791016, "grad_norm": 0.12741616368293762, "learning_rate": 6.3353667433583116e-06, "loss": 0.0224, "step": 135380 }, { "epoch": 2.7055273570201033, "grad_norm": 0.1186675876379013, "learning_rate": 6.330332901999392e-06, "loss": 0.0229, "step": 135390 }, { "epoch": 2.7057271891611046, "grad_norm": 0.09680219739675522, "learning_rate": 6.325300926128441e-06, "loss": 0.0208, "step": 135400 }, { "epoch": 2.7059270213021063, "grad_norm": 0.14865025877952576, "learning_rate": 6.320270815960394e-06, "loss": 0.0248, "step": 135410 }, { "epoch": 2.7061268534431075, "grad_norm": 0.16556863486766815, "learning_rate": 6.315242571710122e-06, "loss": 0.0322, "step": 135420 }, { "epoch": 2.7063266855841093, "grad_norm": 0.14780724048614502, "learning_rate": 6.3102161935924554e-06, "loss": 0.0244, "step": 135430 }, { "epoch": 2.706526517725111, "grad_norm": 0.08784829080104828, "learning_rate": 6.305191681822087e-06, "loss": 0.019, "step": 135440 }, { "epoch": 2.7067263498661127, "grad_norm": 0.11079145967960358, "learning_rate": 6.300169036613657e-06, "loss": 0.0316, "step": 135450 }, { "epoch": 2.706926182007114, "grad_norm": 0.13167086243629456, "learning_rate": 6.295148258181721e-06, "loss": 0.0248, "step": 135460 }, { "epoch": 2.7071260141481157, "grad_norm": 0.10676147043704987, "learning_rate": 6.29012934674077e-06, "loss": 0.0227, "step": 135470 }, { "epoch": 2.707325846289117, "grad_norm": 0.16781160235404968, "learning_rate": 6.285112302505192e-06, "loss": 0.0302, "step": 135480 }, { "epoch": 2.7075256784301187, "grad_norm": 0.12431412190198898, "learning_rate": 6.28009712568931e-06, "loss": 0.0226, "step": 135490 }, { "epoch": 2.7077255105711204, "grad_norm": 0.12046833336353302, "learning_rate": 6.275083816507377e-06, "loss": 0.0243, "step": 135500 }, { "epoch": 2.7079253427121217, "grad_norm": 0.16509570181369781, "learning_rate": 6.270072375173514e-06, "loss": 0.0285, "step": 135510 }, { "epoch": 2.7081251748531234, "grad_norm": 0.18324026465415955, "learning_rate": 6.265062801901839e-06, "loss": 0.0174, "step": 135520 }, { "epoch": 2.7083250069941247, "grad_norm": 0.1459854692220688, "learning_rate": 6.260055096906342e-06, "loss": 0.0344, "step": 135530 }, { "epoch": 2.7085248391351264, "grad_norm": 0.11055092513561249, "learning_rate": 6.25504926040093e-06, "loss": 0.0245, "step": 135540 }, { "epoch": 2.708724671276128, "grad_norm": 0.09544084221124649, "learning_rate": 6.250045292599443e-06, "loss": 0.0229, "step": 135550 }, { "epoch": 2.70892450341713, "grad_norm": 0.12129484862089157, "learning_rate": 6.2450431937156715e-06, "loss": 0.0297, "step": 135560 }, { "epoch": 2.709124335558131, "grad_norm": 0.11794880032539368, "learning_rate": 6.240042963963261e-06, "loss": 0.0291, "step": 135570 }, { "epoch": 2.709324167699133, "grad_norm": 0.10540447384119034, "learning_rate": 6.235044603555829e-06, "loss": 0.0275, "step": 135580 }, { "epoch": 2.709523999840134, "grad_norm": 0.08577419072389603, "learning_rate": 6.230048112706888e-06, "loss": 0.0264, "step": 135590 }, { "epoch": 2.709723831981136, "grad_norm": 0.15501363575458527, "learning_rate": 6.225053491629884e-06, "loss": 0.025, "step": 135600 }, { "epoch": 2.7099236641221376, "grad_norm": 0.12377438694238663, "learning_rate": 6.2200607405381804e-06, "loss": 0.0261, "step": 135610 }, { "epoch": 2.710123496263139, "grad_norm": 0.12351687252521515, "learning_rate": 6.215069859645056e-06, "loss": 0.0229, "step": 135620 }, { "epoch": 2.7103233284041406, "grad_norm": 0.17414917051792145, "learning_rate": 6.210080849163718e-06, "loss": 0.0223, "step": 135630 }, { "epoch": 2.710523160545142, "grad_norm": 0.12921108305454254, "learning_rate": 6.205093709307264e-06, "loss": 0.0228, "step": 135640 }, { "epoch": 2.7107229926861436, "grad_norm": 0.10872350633144379, "learning_rate": 6.200108440288766e-06, "loss": 0.0185, "step": 135650 }, { "epoch": 2.7109228248271453, "grad_norm": 0.10950081795454025, "learning_rate": 6.1951250423211815e-06, "loss": 0.0219, "step": 135660 }, { "epoch": 2.711122656968147, "grad_norm": 0.17670412361621857, "learning_rate": 6.190143515617369e-06, "loss": 0.021, "step": 135670 }, { "epoch": 2.7113224891091483, "grad_norm": 0.09685108810663223, "learning_rate": 6.185163860390147e-06, "loss": 0.0247, "step": 135680 }, { "epoch": 2.71152232125015, "grad_norm": 0.10493403673171997, "learning_rate": 6.180186076852235e-06, "loss": 0.0234, "step": 135690 }, { "epoch": 2.7117221533911513, "grad_norm": 0.1371646374464035, "learning_rate": 6.175210165216272e-06, "loss": 0.0176, "step": 135700 }, { "epoch": 2.711921985532153, "grad_norm": 0.11973720788955688, "learning_rate": 6.170236125694822e-06, "loss": 0.0299, "step": 135710 }, { "epoch": 2.7121218176731547, "grad_norm": 0.11291835457086563, "learning_rate": 6.165263958500367e-06, "loss": 0.0171, "step": 135720 }, { "epoch": 2.712321649814156, "grad_norm": 0.09273751080036163, "learning_rate": 6.160293663845312e-06, "loss": 0.0236, "step": 135730 }, { "epoch": 2.7125214819551577, "grad_norm": 0.1043817549943924, "learning_rate": 6.155325241941972e-06, "loss": 0.0274, "step": 135740 }, { "epoch": 2.712721314096159, "grad_norm": 0.1383177638053894, "learning_rate": 6.150358693002595e-06, "loss": 0.0329, "step": 135750 }, { "epoch": 2.7129211462371607, "grad_norm": 0.11905765533447266, "learning_rate": 6.145394017239336e-06, "loss": 0.0245, "step": 135760 }, { "epoch": 2.7131209783781625, "grad_norm": 0.13374508917331696, "learning_rate": 6.140431214864295e-06, "loss": 0.0263, "step": 135770 }, { "epoch": 2.7133208105191637, "grad_norm": 0.10714667290449142, "learning_rate": 6.13547028608944e-06, "loss": 0.0254, "step": 135780 }, { "epoch": 2.7135206426601655, "grad_norm": 0.17413364350795746, "learning_rate": 6.130511231126718e-06, "loss": 0.0223, "step": 135790 }, { "epoch": 2.713720474801167, "grad_norm": 0.14392085373401642, "learning_rate": 6.125554050187981e-06, "loss": 0.0272, "step": 135800 }, { "epoch": 2.7139203069421685, "grad_norm": 0.16194257140159607, "learning_rate": 6.1205987434849625e-06, "loss": 0.0283, "step": 135810 }, { "epoch": 2.71412013908317, "grad_norm": 0.12653128802776337, "learning_rate": 6.115645311229345e-06, "loss": 0.0212, "step": 135820 }, { "epoch": 2.714319971224172, "grad_norm": 0.09861554950475693, "learning_rate": 6.110693753632768e-06, "loss": 0.0237, "step": 135830 }, { "epoch": 2.714519803365173, "grad_norm": 0.13960610330104828, "learning_rate": 6.1057440709067105e-06, "loss": 0.0273, "step": 135840 }, { "epoch": 2.714719635506175, "grad_norm": 0.12161444872617722, "learning_rate": 6.100796263262632e-06, "loss": 0.0263, "step": 135850 }, { "epoch": 2.714919467647176, "grad_norm": 0.10913161933422089, "learning_rate": 6.09585033091189e-06, "loss": 0.0207, "step": 135860 }, { "epoch": 2.715119299788178, "grad_norm": 0.11155559122562408, "learning_rate": 6.090906274065766e-06, "loss": 0.0243, "step": 135870 }, { "epoch": 2.7153191319291796, "grad_norm": 0.1056688129901886, "learning_rate": 6.0859640929354696e-06, "loss": 0.021, "step": 135880 }, { "epoch": 2.715518964070181, "grad_norm": 0.11006401479244232, "learning_rate": 6.08102378773211e-06, "loss": 0.0282, "step": 135890 }, { "epoch": 2.7157187962111826, "grad_norm": 0.13793547451496124, "learning_rate": 6.076085358666744e-06, "loss": 0.022, "step": 135900 }, { "epoch": 2.715918628352184, "grad_norm": 0.15306496620178223, "learning_rate": 6.071148805950294e-06, "loss": 0.0224, "step": 135910 }, { "epoch": 2.7161184604931856, "grad_norm": 0.1516420841217041, "learning_rate": 6.066214129793685e-06, "loss": 0.0232, "step": 135920 }, { "epoch": 2.7163182926341873, "grad_norm": 0.14569616317749023, "learning_rate": 6.061281330407709e-06, "loss": 0.022, "step": 135930 }, { "epoch": 2.716518124775189, "grad_norm": 0.1180928498506546, "learning_rate": 6.056350408003064e-06, "loss": 0.0203, "step": 135940 }, { "epoch": 2.7167179569161903, "grad_norm": 0.1474905163049698, "learning_rate": 6.051421362790399e-06, "loss": 0.0199, "step": 135950 }, { "epoch": 2.716917789057192, "grad_norm": 0.09947691112756729, "learning_rate": 6.046494194980284e-06, "loss": 0.0261, "step": 135960 }, { "epoch": 2.7171176211981933, "grad_norm": 0.16400650143623352, "learning_rate": 6.04156890478319e-06, "loss": 0.0204, "step": 135970 }, { "epoch": 2.717317453339195, "grad_norm": 0.1835227757692337, "learning_rate": 6.036645492409515e-06, "loss": 0.0274, "step": 135980 }, { "epoch": 2.7175172854801968, "grad_norm": 0.07985598593950272, "learning_rate": 6.031723958069585e-06, "loss": 0.021, "step": 135990 }, { "epoch": 2.717717117621198, "grad_norm": 0.19179701805114746, "learning_rate": 6.026804301973632e-06, "loss": 0.0236, "step": 136000 }, { "epoch": 2.7179169497621998, "grad_norm": 0.21239471435546875, "learning_rate": 6.0218865243318215e-06, "loss": 0.0232, "step": 136010 }, { "epoch": 2.718116781903201, "grad_norm": 0.09536806493997574, "learning_rate": 6.016970625354224e-06, "loss": 0.0272, "step": 136020 }, { "epoch": 2.7183166140442028, "grad_norm": 0.12511767446994781, "learning_rate": 6.012056605250854e-06, "loss": 0.0259, "step": 136030 }, { "epoch": 2.7185164461852045, "grad_norm": 0.11182043701410294, "learning_rate": 6.007144464231607e-06, "loss": 0.0249, "step": 136040 }, { "epoch": 2.718716278326206, "grad_norm": 0.09877883642911911, "learning_rate": 6.002234202506324e-06, "loss": 0.0207, "step": 136050 }, { "epoch": 2.7189161104672075, "grad_norm": 0.10340984165668488, "learning_rate": 5.997325820284788e-06, "loss": 0.0185, "step": 136060 }, { "epoch": 2.719115942608209, "grad_norm": 0.10765454173088074, "learning_rate": 5.992419317776643e-06, "loss": 0.0271, "step": 136070 }, { "epoch": 2.7193157747492105, "grad_norm": 0.10682229697704315, "learning_rate": 5.987514695191504e-06, "loss": 0.0251, "step": 136080 }, { "epoch": 2.719515606890212, "grad_norm": 0.1223270446062088, "learning_rate": 5.982611952738887e-06, "loss": 0.0281, "step": 136090 }, { "epoch": 2.719715439031214, "grad_norm": 0.11886795610189438, "learning_rate": 5.977711090628219e-06, "loss": 0.0265, "step": 136100 }, { "epoch": 2.719915271172215, "grad_norm": 0.1087682694196701, "learning_rate": 5.972812109068865e-06, "loss": 0.0322, "step": 136110 }, { "epoch": 2.720115103313217, "grad_norm": 0.13880394399166107, "learning_rate": 5.967915008270097e-06, "loss": 0.0275, "step": 136120 }, { "epoch": 2.720314935454218, "grad_norm": 0.11008835583925247, "learning_rate": 5.963019788441121e-06, "loss": 0.0262, "step": 136130 }, { "epoch": 2.72051476759522, "grad_norm": 0.15056490898132324, "learning_rate": 5.958126449791018e-06, "loss": 0.0233, "step": 136140 }, { "epoch": 2.7207145997362217, "grad_norm": 0.09037085622549057, "learning_rate": 5.953234992528856e-06, "loss": 0.0242, "step": 136150 }, { "epoch": 2.7209144318772234, "grad_norm": 0.10313906520605087, "learning_rate": 5.948345416863582e-06, "loss": 0.0285, "step": 136160 }, { "epoch": 2.7211142640182246, "grad_norm": 0.12168478220701218, "learning_rate": 5.94345772300407e-06, "loss": 0.0311, "step": 136170 }, { "epoch": 2.7213140961592264, "grad_norm": 0.1264847218990326, "learning_rate": 5.938571911159096e-06, "loss": 0.0296, "step": 136180 }, { "epoch": 2.7215139283002276, "grad_norm": 0.10474690049886703, "learning_rate": 5.933687981537395e-06, "loss": 0.0255, "step": 136190 }, { "epoch": 2.7217137604412294, "grad_norm": 0.16818441450595856, "learning_rate": 5.928805934347597e-06, "loss": 0.0278, "step": 136200 }, { "epoch": 2.721913592582231, "grad_norm": 0.12276097387075424, "learning_rate": 5.923925769798245e-06, "loss": 0.0234, "step": 136210 }, { "epoch": 2.7221134247232324, "grad_norm": 0.09790391474962234, "learning_rate": 5.919047488097801e-06, "loss": 0.0194, "step": 136220 }, { "epoch": 2.722313256864234, "grad_norm": 0.08772806823253632, "learning_rate": 5.9141710894546844e-06, "loss": 0.0233, "step": 136230 }, { "epoch": 2.7225130890052354, "grad_norm": 0.11296004056930542, "learning_rate": 5.909296574077183e-06, "loss": 0.0284, "step": 136240 }, { "epoch": 2.722712921146237, "grad_norm": 0.10536601394414902, "learning_rate": 5.904423942173537e-06, "loss": 0.0207, "step": 136250 }, { "epoch": 2.722912753287239, "grad_norm": 0.1268875151872635, "learning_rate": 5.899553193951896e-06, "loss": 0.0216, "step": 136260 }, { "epoch": 2.7231125854282405, "grad_norm": 0.1279759556055069, "learning_rate": 5.894684329620326e-06, "loss": 0.0207, "step": 136270 }, { "epoch": 2.723312417569242, "grad_norm": 0.09541380405426025, "learning_rate": 5.889817349386811e-06, "loss": 0.0208, "step": 136280 }, { "epoch": 2.7235122497102435, "grad_norm": 0.127680703997612, "learning_rate": 5.884952253459275e-06, "loss": 0.0213, "step": 136290 }, { "epoch": 2.723712081851245, "grad_norm": 0.10498527437448502, "learning_rate": 5.880089042045539e-06, "loss": 0.0252, "step": 136300 }, { "epoch": 2.7239119139922465, "grad_norm": 0.09544091671705246, "learning_rate": 5.875227715353332e-06, "loss": 0.0181, "step": 136310 }, { "epoch": 2.7241117461332482, "grad_norm": 0.125716432929039, "learning_rate": 5.870368273590349e-06, "loss": 0.0283, "step": 136320 }, { "epoch": 2.7243115782742495, "grad_norm": 0.14883816242218018, "learning_rate": 5.865510716964173e-06, "loss": 0.026, "step": 136330 }, { "epoch": 2.7245114104152512, "grad_norm": 0.12472327053546906, "learning_rate": 5.860655045682289e-06, "loss": 0.0344, "step": 136340 }, { "epoch": 2.7247112425562525, "grad_norm": 0.12055444717407227, "learning_rate": 5.855801259952137e-06, "loss": 0.0222, "step": 136350 }, { "epoch": 2.7249110746972542, "grad_norm": 0.1006447821855545, "learning_rate": 5.850949359981062e-06, "loss": 0.0198, "step": 136360 }, { "epoch": 2.725110906838256, "grad_norm": 0.1322149932384491, "learning_rate": 5.846099345976324e-06, "loss": 0.022, "step": 136370 }, { "epoch": 2.7253107389792577, "grad_norm": 0.10754542052745819, "learning_rate": 5.8412512181451105e-06, "loss": 0.0214, "step": 136380 }, { "epoch": 2.725510571120259, "grad_norm": 0.08188996464014053, "learning_rate": 5.83640497669452e-06, "loss": 0.0235, "step": 136390 }, { "epoch": 2.7257104032612607, "grad_norm": 0.11907748878002167, "learning_rate": 5.8315606218315775e-06, "loss": 0.0252, "step": 136400 }, { "epoch": 2.725910235402262, "grad_norm": 0.10572413355112076, "learning_rate": 5.826718153763227e-06, "loss": 0.0261, "step": 136410 }, { "epoch": 2.7261100675432637, "grad_norm": 0.09718228876590729, "learning_rate": 5.821877572696327e-06, "loss": 0.0263, "step": 136420 }, { "epoch": 2.7263098996842654, "grad_norm": 0.12909042835235596, "learning_rate": 5.817038878837666e-06, "loss": 0.0209, "step": 136430 }, { "epoch": 2.7265097318252667, "grad_norm": 0.10404597222805023, "learning_rate": 5.8122020723939355e-06, "loss": 0.0215, "step": 136440 }, { "epoch": 2.7267095639662684, "grad_norm": 0.11551059037446976, "learning_rate": 5.807367153571741e-06, "loss": 0.0228, "step": 136450 }, { "epoch": 2.7269093961072697, "grad_norm": 0.11133092641830444, "learning_rate": 5.802534122577658e-06, "loss": 0.0263, "step": 136460 }, { "epoch": 2.7271092282482714, "grad_norm": 0.13175813853740692, "learning_rate": 5.797702979618113e-06, "loss": 0.0202, "step": 136470 }, { "epoch": 2.727309060389273, "grad_norm": 0.08405880630016327, "learning_rate": 5.792873724899495e-06, "loss": 0.019, "step": 136480 }, { "epoch": 2.727508892530275, "grad_norm": 0.10712632536888123, "learning_rate": 5.788046358628102e-06, "loss": 0.0211, "step": 136490 }, { "epoch": 2.727708724671276, "grad_norm": 0.10737260431051254, "learning_rate": 5.783220881010148e-06, "loss": 0.0214, "step": 136500 }, { "epoch": 2.727908556812278, "grad_norm": 0.1316196769475937, "learning_rate": 5.778397292251769e-06, "loss": 0.0255, "step": 136510 }, { "epoch": 2.728108388953279, "grad_norm": 0.1686384230852127, "learning_rate": 5.773575592559022e-06, "loss": 0.0285, "step": 136520 }, { "epoch": 2.728308221094281, "grad_norm": 0.11638692766427994, "learning_rate": 5.7687557821378856e-06, "loss": 0.0232, "step": 136530 }, { "epoch": 2.7285080532352826, "grad_norm": 0.1589307337999344, "learning_rate": 5.76393786119423e-06, "loss": 0.0309, "step": 136540 }, { "epoch": 2.728707885376284, "grad_norm": 0.11338329315185547, "learning_rate": 5.759121829933894e-06, "loss": 0.0254, "step": 136550 }, { "epoch": 2.7289077175172856, "grad_norm": 0.16062454879283905, "learning_rate": 5.754307688562599e-06, "loss": 0.0412, "step": 136560 }, { "epoch": 2.729107549658287, "grad_norm": 0.11670622229576111, "learning_rate": 5.749495437286007e-06, "loss": 0.0273, "step": 136570 }, { "epoch": 2.7293073817992886, "grad_norm": 0.09064479172229767, "learning_rate": 5.744685076309658e-06, "loss": 0.0182, "step": 136580 }, { "epoch": 2.7295072139402903, "grad_norm": 0.11726831644773483, "learning_rate": 5.739876605839078e-06, "loss": 0.0255, "step": 136590 }, { "epoch": 2.7297070460812916, "grad_norm": 0.15860548615455627, "learning_rate": 5.73507002607967e-06, "loss": 0.0248, "step": 136600 }, { "epoch": 2.7299068782222933, "grad_norm": 0.07784539461135864, "learning_rate": 5.73026533723674e-06, "loss": 0.0309, "step": 136610 }, { "epoch": 2.730106710363295, "grad_norm": 0.089331716299057, "learning_rate": 5.725462539515547e-06, "loss": 0.0216, "step": 136620 }, { "epoch": 2.7303065425042963, "grad_norm": 0.1381990611553192, "learning_rate": 5.720661633121266e-06, "loss": 0.0239, "step": 136630 }, { "epoch": 2.730506374645298, "grad_norm": 0.10932870954275131, "learning_rate": 5.715862618258971e-06, "loss": 0.0212, "step": 136640 }, { "epoch": 2.7307062067862997, "grad_norm": 0.13583983480930328, "learning_rate": 5.711065495133672e-06, "loss": 0.0238, "step": 136650 }, { "epoch": 2.730906038927301, "grad_norm": 0.12322845309972763, "learning_rate": 5.706270263950297e-06, "loss": 0.0222, "step": 136660 }, { "epoch": 2.7311058710683027, "grad_norm": 0.1929735541343689, "learning_rate": 5.701476924913684e-06, "loss": 0.0229, "step": 136670 }, { "epoch": 2.731305703209304, "grad_norm": 0.11347904801368713, "learning_rate": 5.696685478228602e-06, "loss": 0.026, "step": 136680 }, { "epoch": 2.7315055353503057, "grad_norm": 0.11657635122537613, "learning_rate": 5.691895924099722e-06, "loss": 0.0193, "step": 136690 }, { "epoch": 2.7317053674913074, "grad_norm": 0.12407089024782181, "learning_rate": 5.687108262731666e-06, "loss": 0.0279, "step": 136700 }, { "epoch": 2.7319051996323087, "grad_norm": 0.17721012234687805, "learning_rate": 5.682322494328929e-06, "loss": 0.0208, "step": 136710 }, { "epoch": 2.7321050317733104, "grad_norm": 0.15122726559638977, "learning_rate": 5.677538619095951e-06, "loss": 0.0268, "step": 136720 }, { "epoch": 2.7323048639143117, "grad_norm": 0.1291196048259735, "learning_rate": 5.6727566372371195e-06, "loss": 0.0268, "step": 136730 }, { "epoch": 2.7325046960553134, "grad_norm": 0.1198028177022934, "learning_rate": 5.667976548956683e-06, "loss": 0.0248, "step": 136740 }, { "epoch": 2.732704528196315, "grad_norm": 0.10641466081142426, "learning_rate": 5.663198354458849e-06, "loss": 0.0201, "step": 136750 }, { "epoch": 2.732904360337317, "grad_norm": 0.1457359492778778, "learning_rate": 5.658422053947726e-06, "loss": 0.0248, "step": 136760 }, { "epoch": 2.733104192478318, "grad_norm": 0.11152499169111252, "learning_rate": 5.6536476476273635e-06, "loss": 0.029, "step": 136770 }, { "epoch": 2.73330402461932, "grad_norm": 0.08325516432523727, "learning_rate": 5.648875135701698e-06, "loss": 0.0208, "step": 136780 }, { "epoch": 2.733503856760321, "grad_norm": 0.11001627147197723, "learning_rate": 5.644104518374615e-06, "loss": 0.0312, "step": 136790 }, { "epoch": 2.733703688901323, "grad_norm": 0.08131390810012817, "learning_rate": 5.639335795849915e-06, "loss": 0.0228, "step": 136800 }, { "epoch": 2.7339035210423246, "grad_norm": 0.12190526723861694, "learning_rate": 5.6345689683312774e-06, "loss": 0.0327, "step": 136810 }, { "epoch": 2.734103353183326, "grad_norm": 0.0833263099193573, "learning_rate": 5.629804036022357e-06, "loss": 0.0293, "step": 136820 }, { "epoch": 2.7343031853243276, "grad_norm": 0.12872982025146484, "learning_rate": 5.625040999126707e-06, "loss": 0.0216, "step": 136830 }, { "epoch": 2.734503017465329, "grad_norm": 0.08961083739995956, "learning_rate": 5.620279857847782e-06, "loss": 0.0219, "step": 136840 }, { "epoch": 2.7347028496063306, "grad_norm": 0.1386736035346985, "learning_rate": 5.615520612388958e-06, "loss": 0.0223, "step": 136850 }, { "epoch": 2.7349026817473323, "grad_norm": 0.1150851845741272, "learning_rate": 5.6107632629535775e-06, "loss": 0.0259, "step": 136860 }, { "epoch": 2.735102513888334, "grad_norm": 0.09523263573646545, "learning_rate": 5.60600780974484e-06, "loss": 0.0162, "step": 136870 }, { "epoch": 2.7353023460293353, "grad_norm": 0.12483333051204681, "learning_rate": 5.6012542529658875e-06, "loss": 0.0224, "step": 136880 }, { "epoch": 2.735502178170337, "grad_norm": 0.1467016190290451, "learning_rate": 5.596502592819791e-06, "loss": 0.0216, "step": 136890 }, { "epoch": 2.7357020103113383, "grad_norm": 0.0787615031003952, "learning_rate": 5.591752829509539e-06, "loss": 0.0222, "step": 136900 }, { "epoch": 2.73590184245234, "grad_norm": 0.1503332406282425, "learning_rate": 5.587004963238018e-06, "loss": 0.0315, "step": 136910 }, { "epoch": 2.7361016745933417, "grad_norm": 0.11797355860471725, "learning_rate": 5.582258994208056e-06, "loss": 0.0237, "step": 136920 }, { "epoch": 2.736301506734343, "grad_norm": 0.15283328294754028, "learning_rate": 5.577514922622407e-06, "loss": 0.0239, "step": 136930 }, { "epoch": 2.7365013388753447, "grad_norm": 0.1107112318277359, "learning_rate": 5.572772748683691e-06, "loss": 0.022, "step": 136940 }, { "epoch": 2.736701171016346, "grad_norm": 0.1513160765171051, "learning_rate": 5.5680324725945146e-06, "loss": 0.0206, "step": 136950 }, { "epoch": 2.7369010031573477, "grad_norm": 0.113109290599823, "learning_rate": 5.563294094557369e-06, "loss": 0.0181, "step": 136960 }, { "epoch": 2.7371008352983495, "grad_norm": 0.16385096311569214, "learning_rate": 5.558557614774679e-06, "loss": 0.024, "step": 136970 }, { "epoch": 2.737300667439351, "grad_norm": 0.11990049481391907, "learning_rate": 5.553823033448757e-06, "loss": 0.0233, "step": 136980 }, { "epoch": 2.7375004995803525, "grad_norm": 0.10347670316696167, "learning_rate": 5.549090350781849e-06, "loss": 0.023, "step": 136990 }, { "epoch": 2.737700331721354, "grad_norm": 0.09875935316085815, "learning_rate": 5.544359566976165e-06, "loss": 0.0241, "step": 137000 }, { "epoch": 2.7379001638623555, "grad_norm": 0.16392819583415985, "learning_rate": 5.539630682233765e-06, "loss": 0.025, "step": 137010 }, { "epoch": 2.738099996003357, "grad_norm": 0.08214626461267471, "learning_rate": 5.5349036967566655e-06, "loss": 0.0213, "step": 137020 }, { "epoch": 2.738299828144359, "grad_norm": 0.11092860996723175, "learning_rate": 5.530178610746794e-06, "loss": 0.0317, "step": 137030 }, { "epoch": 2.73849966028536, "grad_norm": 0.09457992017269135, "learning_rate": 5.5254554244059955e-06, "loss": 0.0256, "step": 137040 }, { "epoch": 2.738699492426362, "grad_norm": 0.14939269423484802, "learning_rate": 5.520734137936046e-06, "loss": 0.0205, "step": 137050 }, { "epoch": 2.738899324567363, "grad_norm": 0.11729426681995392, "learning_rate": 5.516014751538617e-06, "loss": 0.0175, "step": 137060 }, { "epoch": 2.739099156708365, "grad_norm": 0.16471070051193237, "learning_rate": 5.511297265415333e-06, "loss": 0.0341, "step": 137070 }, { "epoch": 2.7392989888493666, "grad_norm": 0.09331750124692917, "learning_rate": 5.506581679767675e-06, "loss": 0.0214, "step": 137080 }, { "epoch": 2.7394988209903683, "grad_norm": 0.1631479561328888, "learning_rate": 5.501867994797128e-06, "loss": 0.0178, "step": 137090 }, { "epoch": 2.7396986531313696, "grad_norm": 0.10689327120780945, "learning_rate": 5.49715621070504e-06, "loss": 0.0284, "step": 137100 }, { "epoch": 2.7398984852723713, "grad_norm": 0.12735214829444885, "learning_rate": 5.492446327692674e-06, "loss": 0.0315, "step": 137110 }, { "epoch": 2.7400983174133726, "grad_norm": 0.12334708124399185, "learning_rate": 5.4877383459612296e-06, "loss": 0.0274, "step": 137120 }, { "epoch": 2.7402981495543743, "grad_norm": 0.10532332211732864, "learning_rate": 5.483032265711852e-06, "loss": 0.0207, "step": 137130 }, { "epoch": 2.740497981695376, "grad_norm": 0.14282208681106567, "learning_rate": 5.47832808714554e-06, "loss": 0.0228, "step": 137140 }, { "epoch": 2.7406978138363773, "grad_norm": 0.11092568188905716, "learning_rate": 5.4736258104632675e-06, "loss": 0.0187, "step": 137150 }, { "epoch": 2.740897645977379, "grad_norm": 0.09228485822677612, "learning_rate": 5.468925435865907e-06, "loss": 0.0242, "step": 137160 }, { "epoch": 2.7410974781183803, "grad_norm": 0.10310432314872742, "learning_rate": 5.464226963554242e-06, "loss": 0.0235, "step": 137170 }, { "epoch": 2.741297310259382, "grad_norm": 0.10859715193510056, "learning_rate": 5.459530393728984e-06, "loss": 0.0292, "step": 137180 }, { "epoch": 2.741497142400384, "grad_norm": 0.0888262465596199, "learning_rate": 5.454835726590768e-06, "loss": 0.0252, "step": 137190 }, { "epoch": 2.7416969745413855, "grad_norm": 0.10154101252555847, "learning_rate": 5.450142962340143e-06, "loss": 0.0227, "step": 137200 }, { "epoch": 2.741896806682387, "grad_norm": 0.09721643477678299, "learning_rate": 5.445452101177556e-06, "loss": 0.0307, "step": 137210 }, { "epoch": 2.7420966388233885, "grad_norm": 0.12980607151985168, "learning_rate": 5.440763143303413e-06, "loss": 0.0225, "step": 137220 }, { "epoch": 2.7422964709643898, "grad_norm": 0.20463266968727112, "learning_rate": 5.4360760889180186e-06, "loss": 0.0263, "step": 137230 }, { "epoch": 2.7424963031053915, "grad_norm": 0.1278768628835678, "learning_rate": 5.4313909382215755e-06, "loss": 0.0322, "step": 137240 }, { "epoch": 2.742696135246393, "grad_norm": 0.07756961137056351, "learning_rate": 5.42670769141424e-06, "loss": 0.0255, "step": 137250 }, { "epoch": 2.7428959673873945, "grad_norm": 0.13688993453979492, "learning_rate": 5.422026348696069e-06, "loss": 0.0256, "step": 137260 }, { "epoch": 2.743095799528396, "grad_norm": 0.11043393611907959, "learning_rate": 5.417346910267041e-06, "loss": 0.0266, "step": 137270 }, { "epoch": 2.7432956316693975, "grad_norm": 0.13566863536834717, "learning_rate": 5.412669376327051e-06, "loss": 0.0247, "step": 137280 }, { "epoch": 2.743495463810399, "grad_norm": 0.08342614024877548, "learning_rate": 5.407993747075912e-06, "loss": 0.0263, "step": 137290 }, { "epoch": 2.743695295951401, "grad_norm": 0.15145239233970642, "learning_rate": 5.403320022713365e-06, "loss": 0.0295, "step": 137300 }, { "epoch": 2.7438951280924027, "grad_norm": 0.10162938386201859, "learning_rate": 5.39864820343906e-06, "loss": 0.033, "step": 137310 }, { "epoch": 2.744094960233404, "grad_norm": 0.11242030560970306, "learning_rate": 5.393978289452567e-06, "loss": 0.0254, "step": 137320 }, { "epoch": 2.7442947923744057, "grad_norm": 0.11561804264783859, "learning_rate": 5.389310280953391e-06, "loss": 0.0332, "step": 137330 }, { "epoch": 2.744494624515407, "grad_norm": 0.135609969496727, "learning_rate": 5.3846441781409014e-06, "loss": 0.0317, "step": 137340 }, { "epoch": 2.7446944566564087, "grad_norm": 0.10304298996925354, "learning_rate": 5.3799799812144665e-06, "loss": 0.0271, "step": 137350 }, { "epoch": 2.7448942887974104, "grad_norm": 0.1281004548072815, "learning_rate": 5.375317690373316e-06, "loss": 0.023, "step": 137360 }, { "epoch": 2.7450941209384117, "grad_norm": 0.14141151309013367, "learning_rate": 5.370657305816623e-06, "loss": 0.0265, "step": 137370 }, { "epoch": 2.7452939530794134, "grad_norm": 0.12602195143699646, "learning_rate": 5.365998827743457e-06, "loss": 0.0256, "step": 137380 }, { "epoch": 2.7454937852204147, "grad_norm": 0.20597366988658905, "learning_rate": 5.3613422563528136e-06, "loss": 0.0343, "step": 137390 }, { "epoch": 2.7456936173614164, "grad_norm": 0.14765474200248718, "learning_rate": 5.356687591843645e-06, "loss": 0.027, "step": 137400 }, { "epoch": 2.745893449502418, "grad_norm": 0.11250301450490952, "learning_rate": 5.3520348344147655e-06, "loss": 0.0269, "step": 137410 }, { "epoch": 2.7460932816434194, "grad_norm": 0.11352149397134781, "learning_rate": 5.347383984264931e-06, "loss": 0.0227, "step": 137420 }, { "epoch": 2.746293113784421, "grad_norm": 0.10342328995466232, "learning_rate": 5.342735041592822e-06, "loss": 0.0233, "step": 137430 }, { "epoch": 2.7464929459254224, "grad_norm": 0.11494152247905731, "learning_rate": 5.338088006597036e-06, "loss": 0.0247, "step": 137440 }, { "epoch": 2.746692778066424, "grad_norm": 0.14541153609752655, "learning_rate": 5.333442879476086e-06, "loss": 0.0267, "step": 137450 }, { "epoch": 2.746892610207426, "grad_norm": 0.0955064445734024, "learning_rate": 5.328799660428402e-06, "loss": 0.019, "step": 137460 }, { "epoch": 2.7470924423484275, "grad_norm": 0.09632819145917892, "learning_rate": 5.324158349652336e-06, "loss": 0.029, "step": 137470 }, { "epoch": 2.747292274489429, "grad_norm": 0.12091460824012756, "learning_rate": 5.3195189473461416e-06, "loss": 0.0238, "step": 137480 }, { "epoch": 2.7474921066304305, "grad_norm": 0.14179404079914093, "learning_rate": 5.3148814537080205e-06, "loss": 0.0207, "step": 137490 }, { "epoch": 2.747691938771432, "grad_norm": 0.12681549787521362, "learning_rate": 5.310245868936087e-06, "loss": 0.0164, "step": 137500 }, { "epoch": 2.7478917709124335, "grad_norm": 0.14384014904499054, "learning_rate": 5.305612193228343e-06, "loss": 0.0243, "step": 137510 }, { "epoch": 2.7480916030534353, "grad_norm": 0.10366871953010559, "learning_rate": 5.300980426782731e-06, "loss": 0.0328, "step": 137520 }, { "epoch": 2.7482914351944365, "grad_norm": 0.12853503227233887, "learning_rate": 5.296350569797137e-06, "loss": 0.0277, "step": 137530 }, { "epoch": 2.7484912673354382, "grad_norm": 0.1314752697944641, "learning_rate": 5.291722622469314e-06, "loss": 0.0288, "step": 137540 }, { "epoch": 2.7486910994764395, "grad_norm": 0.16475512087345123, "learning_rate": 5.28709658499697e-06, "loss": 0.0193, "step": 137550 }, { "epoch": 2.7488909316174412, "grad_norm": 0.2159651666879654, "learning_rate": 5.28247245757772e-06, "loss": 0.0189, "step": 137560 }, { "epoch": 2.749090763758443, "grad_norm": 0.13163533806800842, "learning_rate": 5.277850240409099e-06, "loss": 0.0297, "step": 137570 }, { "epoch": 2.7492905958994447, "grad_norm": 0.12924642860889435, "learning_rate": 5.273229933688556e-06, "loss": 0.0269, "step": 137580 }, { "epoch": 2.749490428040446, "grad_norm": 0.13555821776390076, "learning_rate": 5.268611537613466e-06, "loss": 0.0229, "step": 137590 }, { "epoch": 2.7496902601814477, "grad_norm": 0.10533164441585541, "learning_rate": 5.263995052381127e-06, "loss": 0.0194, "step": 137600 }, { "epoch": 2.749890092322449, "grad_norm": 0.16073580086231232, "learning_rate": 5.2593804781887136e-06, "loss": 0.0352, "step": 137610 }, { "epoch": 2.7500899244634507, "grad_norm": 0.10257738828659058, "learning_rate": 5.254767815233391e-06, "loss": 0.0223, "step": 137620 }, { "epoch": 2.7502897566044524, "grad_norm": 0.11992961913347244, "learning_rate": 5.250157063712191e-06, "loss": 0.0263, "step": 137630 }, { "epoch": 2.7504895887454537, "grad_norm": 0.17906618118286133, "learning_rate": 5.245548223822067e-06, "loss": 0.037, "step": 137640 }, { "epoch": 2.7506894208864554, "grad_norm": 0.1457982063293457, "learning_rate": 5.240941295759905e-06, "loss": 0.0205, "step": 137650 }, { "epoch": 2.7508892530274567, "grad_norm": 0.12555743753910065, "learning_rate": 5.236336279722509e-06, "loss": 0.0225, "step": 137660 }, { "epoch": 2.7510890851684584, "grad_norm": 0.11070648580789566, "learning_rate": 5.231733175906589e-06, "loss": 0.0237, "step": 137670 }, { "epoch": 2.75128891730946, "grad_norm": 0.10769197344779968, "learning_rate": 5.227131984508788e-06, "loss": 0.022, "step": 137680 }, { "epoch": 2.751488749450462, "grad_norm": 0.08735781162977219, "learning_rate": 5.222532705725658e-06, "loss": 0.0219, "step": 137690 }, { "epoch": 2.751688581591463, "grad_norm": 0.1260465383529663, "learning_rate": 5.217935339753676e-06, "loss": 0.0212, "step": 137700 }, { "epoch": 2.751888413732465, "grad_norm": 0.10796459019184113, "learning_rate": 5.213339886789226e-06, "loss": 0.0223, "step": 137710 }, { "epoch": 2.752088245873466, "grad_norm": 0.10684090852737427, "learning_rate": 5.208746347028615e-06, "loss": 0.0161, "step": 137720 }, { "epoch": 2.752288078014468, "grad_norm": 0.13598421216011047, "learning_rate": 5.2041547206680804e-06, "loss": 0.0222, "step": 137730 }, { "epoch": 2.7524879101554696, "grad_norm": 0.11448965221643448, "learning_rate": 5.199565007903778e-06, "loss": 0.0262, "step": 137740 }, { "epoch": 2.752687742296471, "grad_norm": 0.1326567679643631, "learning_rate": 5.194977208931728e-06, "loss": 0.0235, "step": 137750 }, { "epoch": 2.7528875744374726, "grad_norm": 0.1198529377579689, "learning_rate": 5.190391323947963e-06, "loss": 0.0273, "step": 137760 }, { "epoch": 2.753087406578474, "grad_norm": 0.08004195988178253, "learning_rate": 5.185807353148364e-06, "loss": 0.0218, "step": 137770 }, { "epoch": 2.7532872387194756, "grad_norm": 0.17815087735652924, "learning_rate": 5.181225296728742e-06, "loss": 0.028, "step": 137780 }, { "epoch": 2.7534870708604773, "grad_norm": 0.12952598929405212, "learning_rate": 5.176645154884829e-06, "loss": 0.0189, "step": 137790 }, { "epoch": 2.753686903001479, "grad_norm": 0.11231746524572372, "learning_rate": 5.172066927812314e-06, "loss": 0.0252, "step": 137800 }, { "epoch": 2.7538867351424803, "grad_norm": 0.10832482576370239, "learning_rate": 5.1674906157067385e-06, "loss": 0.022, "step": 137810 }, { "epoch": 2.754086567283482, "grad_norm": 0.10309408605098724, "learning_rate": 5.162916218763603e-06, "loss": 0.0245, "step": 137820 }, { "epoch": 2.7542863994244833, "grad_norm": 0.13213486969470978, "learning_rate": 5.158343737178317e-06, "loss": 0.0181, "step": 137830 }, { "epoch": 2.754486231565485, "grad_norm": 0.12882645428180695, "learning_rate": 5.153773171146209e-06, "loss": 0.0301, "step": 137840 }, { "epoch": 2.7546860637064867, "grad_norm": 0.12435295432806015, "learning_rate": 5.149204520862527e-06, "loss": 0.0247, "step": 137850 }, { "epoch": 2.754885895847488, "grad_norm": 0.11157030612230301, "learning_rate": 5.144637786522438e-06, "loss": 0.0157, "step": 137860 }, { "epoch": 2.7550857279884897, "grad_norm": 0.15026289224624634, "learning_rate": 5.140072968321025e-06, "loss": 0.0209, "step": 137870 }, { "epoch": 2.755285560129491, "grad_norm": 0.1231599971652031, "learning_rate": 5.1355100664532694e-06, "loss": 0.0324, "step": 137880 }, { "epoch": 2.7554853922704927, "grad_norm": 0.09862557053565979, "learning_rate": 5.130949081114111e-06, "loss": 0.0266, "step": 137890 }, { "epoch": 2.7556852244114944, "grad_norm": 0.09778871387243271, "learning_rate": 5.1263900124983945e-06, "loss": 0.0282, "step": 137900 }, { "epoch": 2.755885056552496, "grad_norm": 0.09182623028755188, "learning_rate": 5.121832860800852e-06, "loss": 0.0192, "step": 137910 }, { "epoch": 2.7560848886934974, "grad_norm": 0.1444861739873886, "learning_rate": 5.11727762621616e-06, "loss": 0.0291, "step": 137920 }, { "epoch": 2.756284720834499, "grad_norm": 0.13871149718761444, "learning_rate": 5.112724308938921e-06, "loss": 0.026, "step": 137930 }, { "epoch": 2.7564845529755004, "grad_norm": 0.39147210121154785, "learning_rate": 5.108172909163639e-06, "loss": 0.0204, "step": 137940 }, { "epoch": 2.756684385116502, "grad_norm": 0.18238332867622375, "learning_rate": 5.1036234270847364e-06, "loss": 0.0278, "step": 137950 }, { "epoch": 2.756884217257504, "grad_norm": 0.1803257167339325, "learning_rate": 5.099075862896568e-06, "loss": 0.024, "step": 137960 }, { "epoch": 2.757084049398505, "grad_norm": 0.1891549676656723, "learning_rate": 5.094530216793392e-06, "loss": 0.0262, "step": 137970 }, { "epoch": 2.757283881539507, "grad_norm": 0.13924460113048553, "learning_rate": 5.089986488969389e-06, "loss": 0.0322, "step": 137980 }, { "epoch": 2.757483713680508, "grad_norm": 0.14407530426979065, "learning_rate": 5.0854446796186615e-06, "loss": 0.0246, "step": 137990 }, { "epoch": 2.75768354582151, "grad_norm": 0.13313165307044983, "learning_rate": 5.080904788935242e-06, "loss": 0.027, "step": 138000 }, { "epoch": 2.7578833779625116, "grad_norm": 0.12935714423656464, "learning_rate": 5.076366817113032e-06, "loss": 0.0269, "step": 138010 }, { "epoch": 2.7580832101035133, "grad_norm": 0.09919312596321106, "learning_rate": 5.0718307643459025e-06, "loss": 0.024, "step": 138020 }, { "epoch": 2.7582830422445146, "grad_norm": 0.14314953982830048, "learning_rate": 5.067296630827645e-06, "loss": 0.0235, "step": 138030 }, { "epoch": 2.7584828743855163, "grad_norm": 0.15477819740772247, "learning_rate": 5.062764416751919e-06, "loss": 0.0218, "step": 138040 }, { "epoch": 2.7586827065265176, "grad_norm": 0.18031445145606995, "learning_rate": 5.058234122312344e-06, "loss": 0.0223, "step": 138050 }, { "epoch": 2.7588825386675193, "grad_norm": 0.1289713978767395, "learning_rate": 5.053705747702447e-06, "loss": 0.0277, "step": 138060 }, { "epoch": 2.759082370808521, "grad_norm": 0.21378782391548157, "learning_rate": 5.049179293115674e-06, "loss": 0.021, "step": 138070 }, { "epoch": 2.7592822029495223, "grad_norm": 0.10479564964771271, "learning_rate": 5.044654758745382e-06, "loss": 0.0176, "step": 138080 }, { "epoch": 2.759482035090524, "grad_norm": 0.12904073297977448, "learning_rate": 5.040132144784859e-06, "loss": 0.0256, "step": 138090 }, { "epoch": 2.7596818672315253, "grad_norm": 0.14432868361473083, "learning_rate": 5.035611451427297e-06, "loss": 0.0286, "step": 138100 }, { "epoch": 2.759881699372527, "grad_norm": 0.17002514004707336, "learning_rate": 5.0310926788658006e-06, "loss": 0.0258, "step": 138110 }, { "epoch": 2.7600815315135288, "grad_norm": 0.11613181978464127, "learning_rate": 5.026575827293417e-06, "loss": 0.0309, "step": 138120 }, { "epoch": 2.76028136365453, "grad_norm": 0.12099932134151459, "learning_rate": 5.0220608969031025e-06, "loss": 0.036, "step": 138130 }, { "epoch": 2.7604811957955318, "grad_norm": 0.11023931205272675, "learning_rate": 5.017547887887725e-06, "loss": 0.0284, "step": 138140 }, { "epoch": 2.7606810279365335, "grad_norm": 0.1287626326084137, "learning_rate": 5.013036800440046e-06, "loss": 0.0224, "step": 138150 }, { "epoch": 2.7608808600775347, "grad_norm": 0.12488479912281036, "learning_rate": 5.008527634752802e-06, "loss": 0.028, "step": 138160 }, { "epoch": 2.7610806922185365, "grad_norm": 0.09589605033397675, "learning_rate": 5.004020391018616e-06, "loss": 0.0235, "step": 138170 }, { "epoch": 2.761280524359538, "grad_norm": 0.17641054093837738, "learning_rate": 4.999515069430011e-06, "loss": 0.0235, "step": 138180 }, { "epoch": 2.7614803565005395, "grad_norm": 0.1425647735595703, "learning_rate": 4.9950116701794505e-06, "loss": 0.022, "step": 138190 }, { "epoch": 2.761680188641541, "grad_norm": 0.09800873696804047, "learning_rate": 4.990510193459314e-06, "loss": 0.0285, "step": 138200 }, { "epoch": 2.7618800207825425, "grad_norm": 0.10395842790603638, "learning_rate": 4.986010639461897e-06, "loss": 0.021, "step": 138210 }, { "epoch": 2.762079852923544, "grad_norm": 0.10777410864830017, "learning_rate": 4.9815130083794126e-06, "loss": 0.0252, "step": 138220 }, { "epoch": 2.762279685064546, "grad_norm": 0.1285189986228943, "learning_rate": 4.977017300403991e-06, "loss": 0.032, "step": 138230 }, { "epoch": 2.762479517205547, "grad_norm": 0.12828603386878967, "learning_rate": 4.9725235157276836e-06, "loss": 0.0278, "step": 138240 }, { "epoch": 2.762679349346549, "grad_norm": 0.10532054305076599, "learning_rate": 4.968031654542449e-06, "loss": 0.0237, "step": 138250 }, { "epoch": 2.76287918148755, "grad_norm": 0.13349846005439758, "learning_rate": 4.963541717040177e-06, "loss": 0.0208, "step": 138260 }, { "epoch": 2.763079013628552, "grad_norm": 0.12980853021144867, "learning_rate": 4.95905370341268e-06, "loss": 0.0225, "step": 138270 }, { "epoch": 2.7632788457695536, "grad_norm": 0.17199532687664032, "learning_rate": 4.954567613851657e-06, "loss": 0.024, "step": 138280 }, { "epoch": 2.7634786779105553, "grad_norm": 0.1482962816953659, "learning_rate": 4.950083448548742e-06, "loss": 0.0292, "step": 138290 }, { "epoch": 2.7636785100515566, "grad_norm": 0.15142084658145905, "learning_rate": 4.945601207695522e-06, "loss": 0.0275, "step": 138300 }, { "epoch": 2.7638783421925583, "grad_norm": 0.12412964552640915, "learning_rate": 4.941120891483442e-06, "loss": 0.0242, "step": 138310 }, { "epoch": 2.7640781743335596, "grad_norm": 0.09396775811910629, "learning_rate": 4.9366425001038995e-06, "loss": 0.0178, "step": 138320 }, { "epoch": 2.7642780064745613, "grad_norm": 0.09034296125173569, "learning_rate": 4.932166033748209e-06, "loss": 0.0213, "step": 138330 }, { "epoch": 2.764477838615563, "grad_norm": 0.10277533531188965, "learning_rate": 4.927691492607589e-06, "loss": 0.0348, "step": 138340 }, { "epoch": 2.7646776707565643, "grad_norm": 0.1327337920665741, "learning_rate": 4.923218876873193e-06, "loss": 0.0273, "step": 138350 }, { "epoch": 2.764877502897566, "grad_norm": 0.15693043172359467, "learning_rate": 4.918748186736078e-06, "loss": 0.0398, "step": 138360 }, { "epoch": 2.7650773350385673, "grad_norm": 0.1849086582660675, "learning_rate": 4.91427942238723e-06, "loss": 0.0225, "step": 138370 }, { "epoch": 2.765277167179569, "grad_norm": 0.0892951563000679, "learning_rate": 4.90981258401752e-06, "loss": 0.0178, "step": 138380 }, { "epoch": 2.765476999320571, "grad_norm": 0.14821192622184753, "learning_rate": 4.905347671817795e-06, "loss": 0.0351, "step": 138390 }, { "epoch": 2.7656768314615725, "grad_norm": 0.17405058443546295, "learning_rate": 4.900884685978785e-06, "loss": 0.0282, "step": 138400 }, { "epoch": 2.765876663602574, "grad_norm": 0.10640706866979599, "learning_rate": 4.8964236266911215e-06, "loss": 0.0231, "step": 138410 }, { "epoch": 2.7660764957435755, "grad_norm": 0.15667827427387238, "learning_rate": 4.891964494145374e-06, "loss": 0.0246, "step": 138420 }, { "epoch": 2.766276327884577, "grad_norm": 0.11196272075176239, "learning_rate": 4.887507288532051e-06, "loss": 0.0257, "step": 138430 }, { "epoch": 2.7664761600255785, "grad_norm": 0.1324184685945511, "learning_rate": 4.8830520100415386e-06, "loss": 0.0225, "step": 138440 }, { "epoch": 2.7666759921665802, "grad_norm": 0.11048085242509842, "learning_rate": 4.8785986588641584e-06, "loss": 0.0254, "step": 138450 }, { "epoch": 2.7668758243075815, "grad_norm": 0.11305812001228333, "learning_rate": 4.874147235190157e-06, "loss": 0.022, "step": 138460 }, { "epoch": 2.767075656448583, "grad_norm": 0.0997561439871788, "learning_rate": 4.8696977392096835e-06, "loss": 0.0244, "step": 138470 }, { "epoch": 2.7672754885895845, "grad_norm": 0.09668620675802231, "learning_rate": 4.865250171112818e-06, "loss": 0.0218, "step": 138480 }, { "epoch": 2.767475320730586, "grad_norm": 0.08885156363248825, "learning_rate": 4.860804531089552e-06, "loss": 0.0193, "step": 138490 }, { "epoch": 2.767675152871588, "grad_norm": 0.0996592715382576, "learning_rate": 4.8563608193297974e-06, "loss": 0.0183, "step": 138500 }, { "epoch": 2.7678749850125897, "grad_norm": 0.09775334596633911, "learning_rate": 4.851919036023361e-06, "loss": 0.028, "step": 138510 }, { "epoch": 2.768074817153591, "grad_norm": 0.12412670999765396, "learning_rate": 4.84747918136001e-06, "loss": 0.0235, "step": 138520 }, { "epoch": 2.7682746492945927, "grad_norm": 0.13387849926948547, "learning_rate": 4.843041255529407e-06, "loss": 0.0223, "step": 138530 }, { "epoch": 2.768474481435594, "grad_norm": 0.15286126732826233, "learning_rate": 4.838605258721129e-06, "loss": 0.022, "step": 138540 }, { "epoch": 2.7686743135765957, "grad_norm": 0.11473348736763, "learning_rate": 4.834171191124653e-06, "loss": 0.0269, "step": 138550 }, { "epoch": 2.7688741457175974, "grad_norm": 0.16185644268989563, "learning_rate": 4.829739052929422e-06, "loss": 0.0257, "step": 138560 }, { "epoch": 2.7690739778585987, "grad_norm": 0.1791391670703888, "learning_rate": 4.8253088443247665e-06, "loss": 0.026, "step": 138570 }, { "epoch": 2.7692738099996004, "grad_norm": 0.1535450667142868, "learning_rate": 4.820880565499919e-06, "loss": 0.0317, "step": 138580 }, { "epoch": 2.7694736421406017, "grad_norm": 0.1809602677822113, "learning_rate": 4.816454216644056e-06, "loss": 0.0205, "step": 138590 }, { "epoch": 2.7696734742816034, "grad_norm": 0.13555563986301422, "learning_rate": 4.812029797946266e-06, "loss": 0.022, "step": 138600 }, { "epoch": 2.769873306422605, "grad_norm": 0.11103430390357971, "learning_rate": 4.807607309595552e-06, "loss": 0.0235, "step": 138610 }, { "epoch": 2.770073138563607, "grad_norm": 0.10543935000896454, "learning_rate": 4.8031867517808305e-06, "loss": 0.0254, "step": 138620 }, { "epoch": 2.770272970704608, "grad_norm": 0.09406281262636185, "learning_rate": 4.798768124690939e-06, "loss": 0.0221, "step": 138630 }, { "epoch": 2.77047280284561, "grad_norm": 0.14502227306365967, "learning_rate": 4.794351428514643e-06, "loss": 0.0234, "step": 138640 }, { "epoch": 2.770672634986611, "grad_norm": 0.11699213832616806, "learning_rate": 4.789936663440603e-06, "loss": 0.019, "step": 138650 }, { "epoch": 2.770872467127613, "grad_norm": 0.08273740857839584, "learning_rate": 4.785523829657412e-06, "loss": 0.0222, "step": 138660 }, { "epoch": 2.7710722992686145, "grad_norm": 0.14728371798992157, "learning_rate": 4.781112927353598e-06, "loss": 0.0243, "step": 138670 }, { "epoch": 2.771272131409616, "grad_norm": 0.1640661656856537, "learning_rate": 4.77670395671756e-06, "loss": 0.027, "step": 138680 }, { "epoch": 2.7714719635506175, "grad_norm": 0.1528010219335556, "learning_rate": 4.772296917937636e-06, "loss": 0.0246, "step": 138690 }, { "epoch": 2.771671795691619, "grad_norm": 0.16195829212665558, "learning_rate": 4.767891811202124e-06, "loss": 0.0263, "step": 138700 }, { "epoch": 2.7718716278326205, "grad_norm": 0.10598541051149368, "learning_rate": 4.763488636699171e-06, "loss": 0.0193, "step": 138710 }, { "epoch": 2.7720714599736223, "grad_norm": 0.2511017620563507, "learning_rate": 4.759087394616879e-06, "loss": 0.0247, "step": 138720 }, { "epoch": 2.772271292114624, "grad_norm": 0.12133422493934631, "learning_rate": 4.754688085143267e-06, "loss": 0.0215, "step": 138730 }, { "epoch": 2.7724711242556253, "grad_norm": 0.17916221916675568, "learning_rate": 4.750290708466259e-06, "loss": 0.0251, "step": 138740 }, { "epoch": 2.772670956396627, "grad_norm": 0.12281982600688934, "learning_rate": 4.745895264773703e-06, "loss": 0.0195, "step": 138750 }, { "epoch": 2.7728707885376283, "grad_norm": 0.1207389160990715, "learning_rate": 4.741501754253374e-06, "loss": 0.0216, "step": 138760 }, { "epoch": 2.77307062067863, "grad_norm": 0.13034972548484802, "learning_rate": 4.737110177092952e-06, "loss": 0.0292, "step": 138770 }, { "epoch": 2.7732704528196317, "grad_norm": 0.0832262858748436, "learning_rate": 4.732720533480012e-06, "loss": 0.0187, "step": 138780 }, { "epoch": 2.773470284960633, "grad_norm": 0.11314869672060013, "learning_rate": 4.728332823602105e-06, "loss": 0.0248, "step": 138790 }, { "epoch": 2.7736701171016347, "grad_norm": 0.11039743572473526, "learning_rate": 4.7239470476466655e-06, "loss": 0.0218, "step": 138800 }, { "epoch": 2.773869949242636, "grad_norm": 0.1217556968331337, "learning_rate": 4.7195632058010195e-06, "loss": 0.0165, "step": 138810 }, { "epoch": 2.7740697813836377, "grad_norm": 0.08621668070554733, "learning_rate": 4.715181298252441e-06, "loss": 0.0236, "step": 138820 }, { "epoch": 2.7742696135246394, "grad_norm": 0.12026449292898178, "learning_rate": 4.710801325188147e-06, "loss": 0.0223, "step": 138830 }, { "epoch": 2.774469445665641, "grad_norm": 0.1252751499414444, "learning_rate": 4.7064232867952084e-06, "loss": 0.0243, "step": 138840 }, { "epoch": 2.7746692778066424, "grad_norm": 0.1341872662305832, "learning_rate": 4.702047183260666e-06, "loss": 0.029, "step": 138850 }, { "epoch": 2.774869109947644, "grad_norm": 0.15200059115886688, "learning_rate": 4.697673014771448e-06, "loss": 0.0233, "step": 138860 }, { "epoch": 2.7750689420886454, "grad_norm": 0.1986113041639328, "learning_rate": 4.693300781514421e-06, "loss": 0.0206, "step": 138870 }, { "epoch": 2.775268774229647, "grad_norm": 0.14172832667827606, "learning_rate": 4.688930483676346e-06, "loss": 0.0231, "step": 138880 }, { "epoch": 2.775468606370649, "grad_norm": 0.08373916894197464, "learning_rate": 4.684562121443925e-06, "loss": 0.0173, "step": 138890 }, { "epoch": 2.77566843851165, "grad_norm": 0.12368245422840118, "learning_rate": 4.680195695003775e-06, "loss": 0.028, "step": 138900 }, { "epoch": 2.775868270652652, "grad_norm": 0.16887541115283966, "learning_rate": 4.675831204542391e-06, "loss": 0.0324, "step": 138910 }, { "epoch": 2.776068102793653, "grad_norm": 0.10347096621990204, "learning_rate": 4.6714686502462414e-06, "loss": 0.0183, "step": 138920 }, { "epoch": 2.776267934934655, "grad_norm": 0.09042053669691086, "learning_rate": 4.667108032301681e-06, "loss": 0.0321, "step": 138930 }, { "epoch": 2.7764677670756566, "grad_norm": 0.1400481015443802, "learning_rate": 4.662749350894996e-06, "loss": 0.0314, "step": 138940 }, { "epoch": 2.776667599216658, "grad_norm": 0.11756515502929688, "learning_rate": 4.658392606212359e-06, "loss": 0.0219, "step": 138950 }, { "epoch": 2.7768674313576596, "grad_norm": 0.1695026308298111, "learning_rate": 4.654037798439892e-06, "loss": 0.0228, "step": 138960 }, { "epoch": 2.7770672634986613, "grad_norm": 0.14315477013587952, "learning_rate": 4.649684927763642e-06, "loss": 0.0286, "step": 138970 }, { "epoch": 2.7772670956396626, "grad_norm": 0.15879376232624054, "learning_rate": 4.645333994369527e-06, "loss": 0.0262, "step": 138980 }, { "epoch": 2.7774669277806643, "grad_norm": 0.14818987250328064, "learning_rate": 4.640984998443432e-06, "loss": 0.0211, "step": 138990 }, { "epoch": 2.777666759921666, "grad_norm": 0.14453037083148956, "learning_rate": 4.636637940171123e-06, "loss": 0.0344, "step": 139000 }, { "epoch": 2.7778665920626673, "grad_norm": 0.06822619587182999, "learning_rate": 4.632292819738315e-06, "loss": 0.0199, "step": 139010 }, { "epoch": 2.778066424203669, "grad_norm": 0.08631286770105362, "learning_rate": 4.6279496373306085e-06, "loss": 0.0329, "step": 139020 }, { "epoch": 2.7782662563446703, "grad_norm": 0.12713943421840668, "learning_rate": 4.6236083931335436e-06, "loss": 0.03, "step": 139030 }, { "epoch": 2.778466088485672, "grad_norm": 0.2228005975484848, "learning_rate": 4.619269087332578e-06, "loss": 0.0269, "step": 139040 }, { "epoch": 2.7786659206266737, "grad_norm": 0.15165188908576965, "learning_rate": 4.614931720113053e-06, "loss": 0.0165, "step": 139050 }, { "epoch": 2.778865752767675, "grad_norm": 0.13731178641319275, "learning_rate": 4.61059629166028e-06, "loss": 0.0284, "step": 139060 }, { "epoch": 2.7790655849086767, "grad_norm": 0.11018839478492737, "learning_rate": 4.6062628021594554e-06, "loss": 0.0204, "step": 139070 }, { "epoch": 2.779265417049678, "grad_norm": 0.07232848554849625, "learning_rate": 4.601931251795688e-06, "loss": 0.0253, "step": 139080 }, { "epoch": 2.7794652491906797, "grad_norm": 0.1294383853673935, "learning_rate": 4.5976016407540065e-06, "loss": 0.0256, "step": 139090 }, { "epoch": 2.7796650813316814, "grad_norm": 0.12808451056480408, "learning_rate": 4.593273969219397e-06, "loss": 0.0247, "step": 139100 }, { "epoch": 2.779864913472683, "grad_norm": 0.10311602801084518, "learning_rate": 4.588948237376695e-06, "loss": 0.0288, "step": 139110 }, { "epoch": 2.7800647456136844, "grad_norm": 0.069104865193367, "learning_rate": 4.5846244454107065e-06, "loss": 0.0204, "step": 139120 }, { "epoch": 2.780264577754686, "grad_norm": 0.138897106051445, "learning_rate": 4.5803025935061306e-06, "loss": 0.0252, "step": 139130 }, { "epoch": 2.7804644098956874, "grad_norm": 0.12838122248649597, "learning_rate": 4.5759826818475844e-06, "loss": 0.0287, "step": 139140 }, { "epoch": 2.780664242036689, "grad_norm": 0.09734787046909332, "learning_rate": 4.571664710619616e-06, "loss": 0.0276, "step": 139150 }, { "epoch": 2.780864074177691, "grad_norm": 0.2865782380104065, "learning_rate": 4.567348680006672e-06, "loss": 0.0253, "step": 139160 }, { "epoch": 2.781063906318692, "grad_norm": 0.10892672836780548, "learning_rate": 4.5630345901931435e-06, "loss": 0.0243, "step": 139170 }, { "epoch": 2.781263738459694, "grad_norm": 0.14380952715873718, "learning_rate": 4.558722441363289e-06, "loss": 0.0243, "step": 139180 }, { "epoch": 2.781463570600695, "grad_norm": 0.0950310006737709, "learning_rate": 4.554412233701344e-06, "loss": 0.0254, "step": 139190 }, { "epoch": 2.781663402741697, "grad_norm": 0.12504208087921143, "learning_rate": 4.550103967391428e-06, "loss": 0.0229, "step": 139200 }, { "epoch": 2.7818632348826986, "grad_norm": 0.09736302495002747, "learning_rate": 4.545797642617566e-06, "loss": 0.0306, "step": 139210 }, { "epoch": 2.7820630670237003, "grad_norm": 0.0877082496881485, "learning_rate": 4.541493259563728e-06, "loss": 0.0281, "step": 139220 }, { "epoch": 2.7822628991647016, "grad_norm": 0.07595549523830414, "learning_rate": 4.537190818413789e-06, "loss": 0.0237, "step": 139230 }, { "epoch": 2.7824627313057033, "grad_norm": 0.1190401241183281, "learning_rate": 4.53289031935154e-06, "loss": 0.0285, "step": 139240 }, { "epoch": 2.7826625634467046, "grad_norm": 0.1219678744673729, "learning_rate": 4.5285917625606915e-06, "loss": 0.0219, "step": 139250 }, { "epoch": 2.7828623955877063, "grad_norm": 0.11846514046192169, "learning_rate": 4.524295148224867e-06, "loss": 0.0265, "step": 139260 }, { "epoch": 2.783062227728708, "grad_norm": 0.0959438756108284, "learning_rate": 4.520000476527614e-06, "loss": 0.0147, "step": 139270 }, { "epoch": 2.7832620598697093, "grad_norm": 0.11715856194496155, "learning_rate": 4.5157077476523925e-06, "loss": 0.0274, "step": 139280 }, { "epoch": 2.783461892010711, "grad_norm": 0.1483439952135086, "learning_rate": 4.511416961782578e-06, "loss": 0.0203, "step": 139290 }, { "epoch": 2.7836617241517123, "grad_norm": 0.120354562997818, "learning_rate": 4.5071281191014725e-06, "loss": 0.0241, "step": 139300 }, { "epoch": 2.783861556292714, "grad_norm": 0.12649989128112793, "learning_rate": 4.50284121979227e-06, "loss": 0.0183, "step": 139310 }, { "epoch": 2.7840613884337158, "grad_norm": 0.10179586708545685, "learning_rate": 4.4985562640381065e-06, "loss": 0.0264, "step": 139320 }, { "epoch": 2.7842612205747175, "grad_norm": 0.08702807128429413, "learning_rate": 4.494273252022036e-06, "loss": 0.0229, "step": 139330 }, { "epoch": 2.7844610527157188, "grad_norm": 0.11622799932956696, "learning_rate": 4.489992183927022e-06, "loss": 0.0249, "step": 139340 }, { "epoch": 2.7846608848567205, "grad_norm": 0.10667939484119415, "learning_rate": 4.485713059935937e-06, "loss": 0.0274, "step": 139350 }, { "epoch": 2.7848607169977218, "grad_norm": 0.10138554126024246, "learning_rate": 4.481435880231561e-06, "loss": 0.0168, "step": 139360 }, { "epoch": 2.7850605491387235, "grad_norm": 0.12781518697738647, "learning_rate": 4.477160644996642e-06, "loss": 0.0224, "step": 139370 }, { "epoch": 2.785260381279725, "grad_norm": 0.10223890095949173, "learning_rate": 4.472887354413785e-06, "loss": 0.0275, "step": 139380 }, { "epoch": 2.7854602134207265, "grad_norm": 0.1088998094201088, "learning_rate": 4.468616008665544e-06, "loss": 0.0224, "step": 139390 }, { "epoch": 2.785660045561728, "grad_norm": 0.11142022907733917, "learning_rate": 4.464346607934383e-06, "loss": 0.0239, "step": 139400 }, { "epoch": 2.7858598777027295, "grad_norm": 0.11161354184150696, "learning_rate": 4.460079152402685e-06, "loss": 0.0241, "step": 139410 }, { "epoch": 2.786059709843731, "grad_norm": 0.11080773919820786, "learning_rate": 4.455813642252749e-06, "loss": 0.0259, "step": 139420 }, { "epoch": 2.786259541984733, "grad_norm": 0.14261086285114288, "learning_rate": 4.4515500776667826e-06, "loss": 0.0254, "step": 139430 }, { "epoch": 2.7864593741257346, "grad_norm": 0.09229959547519684, "learning_rate": 4.447288458826931e-06, "loss": 0.0249, "step": 139440 }, { "epoch": 2.786659206266736, "grad_norm": 0.1672307252883911, "learning_rate": 4.443028785915215e-06, "loss": 0.0235, "step": 139450 }, { "epoch": 2.7868590384077376, "grad_norm": 0.1154249757528305, "learning_rate": 4.438771059113628e-06, "loss": 0.0254, "step": 139460 }, { "epoch": 2.787058870548739, "grad_norm": 0.11523047834634781, "learning_rate": 4.434515278604051e-06, "loss": 0.0224, "step": 139470 }, { "epoch": 2.7872587026897406, "grad_norm": 0.11252528429031372, "learning_rate": 4.430261444568268e-06, "loss": 0.0184, "step": 139480 }, { "epoch": 2.7874585348307424, "grad_norm": 0.1520465910434723, "learning_rate": 4.426009557187999e-06, "loss": 0.0191, "step": 139490 }, { "epoch": 2.7876583669717436, "grad_norm": 0.12655936181545258, "learning_rate": 4.421759616644877e-06, "loss": 0.026, "step": 139500 }, { "epoch": 2.7878581991127453, "grad_norm": 0.10606712847948074, "learning_rate": 4.4175116231204625e-06, "loss": 0.025, "step": 139510 }, { "epoch": 2.7880580312537466, "grad_norm": 0.16174153983592987, "learning_rate": 4.413265576796205e-06, "loss": 0.0286, "step": 139520 }, { "epoch": 2.7882578633947483, "grad_norm": 0.11041764169931412, "learning_rate": 4.409021477853503e-06, "loss": 0.0257, "step": 139530 }, { "epoch": 2.78845769553575, "grad_norm": 0.17969314754009247, "learning_rate": 4.404779326473646e-06, "loss": 0.0256, "step": 139540 }, { "epoch": 2.788657527676752, "grad_norm": 0.1291903555393219, "learning_rate": 4.40053912283786e-06, "loss": 0.0237, "step": 139550 }, { "epoch": 2.788857359817753, "grad_norm": 0.09790248423814774, "learning_rate": 4.396300867127268e-06, "loss": 0.0225, "step": 139560 }, { "epoch": 2.789057191958755, "grad_norm": 0.1380721926689148, "learning_rate": 4.39206455952294e-06, "loss": 0.0244, "step": 139570 }, { "epoch": 2.789257024099756, "grad_norm": 0.09787099808454514, "learning_rate": 4.387830200205817e-06, "loss": 0.0243, "step": 139580 }, { "epoch": 2.789456856240758, "grad_norm": 0.14680851995944977, "learning_rate": 4.383597789356792e-06, "loss": 0.0218, "step": 139590 }, { "epoch": 2.7896566883817595, "grad_norm": 0.12223386764526367, "learning_rate": 4.3793673271566805e-06, "loss": 0.0257, "step": 139600 }, { "epoch": 2.789856520522761, "grad_norm": 0.07805464416742325, "learning_rate": 4.375138813786184e-06, "loss": 0.0201, "step": 139610 }, { "epoch": 2.7900563526637625, "grad_norm": 0.11282540112733841, "learning_rate": 4.3709122494259405e-06, "loss": 0.0219, "step": 139620 }, { "epoch": 2.790256184804764, "grad_norm": 0.1805679202079773, "learning_rate": 4.366687634256505e-06, "loss": 0.0332, "step": 139630 }, { "epoch": 2.7904560169457655, "grad_norm": 0.10236471891403198, "learning_rate": 4.362464968458341e-06, "loss": 0.0265, "step": 139640 }, { "epoch": 2.7906558490867672, "grad_norm": 0.12536180019378662, "learning_rate": 4.358244252211835e-06, "loss": 0.0255, "step": 139650 }, { "epoch": 2.790855681227769, "grad_norm": 0.1290438026189804, "learning_rate": 4.354025485697289e-06, "loss": 0.0203, "step": 139660 }, { "epoch": 2.7910555133687702, "grad_norm": 0.11389383673667908, "learning_rate": 4.349808669094924e-06, "loss": 0.0307, "step": 139670 }, { "epoch": 2.791255345509772, "grad_norm": 0.13534511625766754, "learning_rate": 4.345593802584857e-06, "loss": 0.0241, "step": 139680 }, { "epoch": 2.7914551776507732, "grad_norm": 0.12037075310945511, "learning_rate": 4.341380886347163e-06, "loss": 0.0227, "step": 139690 }, { "epoch": 2.791655009791775, "grad_norm": 0.1046801209449768, "learning_rate": 4.337169920561806e-06, "loss": 0.0247, "step": 139700 }, { "epoch": 2.7918548419327767, "grad_norm": 0.11484693735837936, "learning_rate": 4.332960905408656e-06, "loss": 0.0263, "step": 139710 }, { "epoch": 2.792054674073778, "grad_norm": 0.11553294956684113, "learning_rate": 4.3287538410675125e-06, "loss": 0.022, "step": 139720 }, { "epoch": 2.7922545062147797, "grad_norm": 0.13940611481666565, "learning_rate": 4.324548727718114e-06, "loss": 0.0359, "step": 139730 }, { "epoch": 2.792454338355781, "grad_norm": 0.11784309148788452, "learning_rate": 4.320345565540096e-06, "loss": 0.0225, "step": 139740 }, { "epoch": 2.7926541704967827, "grad_norm": 0.1404307782649994, "learning_rate": 4.316144354712992e-06, "loss": 0.0169, "step": 139750 }, { "epoch": 2.7928540026377844, "grad_norm": 0.13574539124965668, "learning_rate": 4.311945095416264e-06, "loss": 0.0277, "step": 139760 }, { "epoch": 2.7930538347787857, "grad_norm": 0.1515861302614212, "learning_rate": 4.30774778782933e-06, "loss": 0.0292, "step": 139770 }, { "epoch": 2.7932536669197874, "grad_norm": 0.1960822194814682, "learning_rate": 4.303552432131464e-06, "loss": 0.0183, "step": 139780 }, { "epoch": 2.793453499060789, "grad_norm": 0.13005094230175018, "learning_rate": 4.299359028501887e-06, "loss": 0.0321, "step": 139790 }, { "epoch": 2.7936533312017904, "grad_norm": 0.11441910266876221, "learning_rate": 4.295167577119741e-06, "loss": 0.0217, "step": 139800 }, { "epoch": 2.793853163342792, "grad_norm": 0.07496161758899689, "learning_rate": 4.290978078164071e-06, "loss": 0.0168, "step": 139810 }, { "epoch": 2.794052995483794, "grad_norm": 0.12109774351119995, "learning_rate": 4.28679053181385e-06, "loss": 0.0289, "step": 139820 }, { "epoch": 2.794252827624795, "grad_norm": 0.08729954808950424, "learning_rate": 4.282604938247958e-06, "loss": 0.0257, "step": 139830 }, { "epoch": 2.794452659765797, "grad_norm": 0.1833372712135315, "learning_rate": 4.278421297645208e-06, "loss": 0.0217, "step": 139840 }, { "epoch": 2.794652491906798, "grad_norm": 0.18762680888175964, "learning_rate": 4.274239610184289e-06, "loss": 0.0221, "step": 139850 }, { "epoch": 2.7948523240478, "grad_norm": 0.12273762375116348, "learning_rate": 4.270059876043858e-06, "loss": 0.0283, "step": 139860 }, { "epoch": 2.7950521561888015, "grad_norm": 0.1594916731119156, "learning_rate": 4.265882095402474e-06, "loss": 0.0246, "step": 139870 }, { "epoch": 2.795251988329803, "grad_norm": 0.14053578674793243, "learning_rate": 4.2617062684385815e-06, "loss": 0.0268, "step": 139880 }, { "epoch": 2.7954518204708045, "grad_norm": 0.09450677782297134, "learning_rate": 4.257532395330571e-06, "loss": 0.026, "step": 139890 }, { "epoch": 2.795651652611806, "grad_norm": 0.10159463435411453, "learning_rate": 4.253360476256746e-06, "loss": 0.0158, "step": 139900 }, { "epoch": 2.7958514847528075, "grad_norm": 0.13998688757419586, "learning_rate": 4.249190511395318e-06, "loss": 0.0233, "step": 139910 }, { "epoch": 2.7960513168938093, "grad_norm": 0.1363804042339325, "learning_rate": 4.245022500924428e-06, "loss": 0.021, "step": 139920 }, { "epoch": 2.796251149034811, "grad_norm": 0.1132432147860527, "learning_rate": 4.2408564450221234e-06, "loss": 0.0246, "step": 139930 }, { "epoch": 2.7964509811758123, "grad_norm": 0.14557702839374542, "learning_rate": 4.236692343866372e-06, "loss": 0.0232, "step": 139940 }, { "epoch": 2.796650813316814, "grad_norm": 0.14951664209365845, "learning_rate": 4.2325301976350484e-06, "loss": 0.0248, "step": 139950 }, { "epoch": 2.7968506454578153, "grad_norm": 0.10538702458143234, "learning_rate": 4.228370006505961e-06, "loss": 0.0231, "step": 139960 }, { "epoch": 2.797050477598817, "grad_norm": 0.10431618988513947, "learning_rate": 4.224211770656828e-06, "loss": 0.0282, "step": 139970 }, { "epoch": 2.7972503097398187, "grad_norm": 0.1522417515516281, "learning_rate": 4.2200554902652625e-06, "loss": 0.0213, "step": 139980 }, { "epoch": 2.79745014188082, "grad_norm": 0.18400351703166962, "learning_rate": 4.215901165508823e-06, "loss": 0.0257, "step": 139990 }, { "epoch": 2.7976499740218217, "grad_norm": 0.11985348165035248, "learning_rate": 4.211748796564996e-06, "loss": 0.0285, "step": 140000 }, { "epoch": 2.797849806162823, "grad_norm": 0.11720967292785645, "learning_rate": 4.207598383611133e-06, "loss": 0.0235, "step": 140010 }, { "epoch": 2.7980496383038247, "grad_norm": 0.11634858697652817, "learning_rate": 4.203449926824543e-06, "loss": 0.0248, "step": 140020 }, { "epoch": 2.7982494704448264, "grad_norm": 0.11255621165037155, "learning_rate": 4.199303426382445e-06, "loss": 0.0186, "step": 140030 }, { "epoch": 2.798449302585828, "grad_norm": 0.1454826146364212, "learning_rate": 4.195158882461964e-06, "loss": 0.0179, "step": 140040 }, { "epoch": 2.7986491347268294, "grad_norm": 0.12031717598438263, "learning_rate": 4.191016295240146e-06, "loss": 0.0228, "step": 140050 }, { "epoch": 2.798848966867831, "grad_norm": 0.1386386603116989, "learning_rate": 4.186875664893963e-06, "loss": 0.0253, "step": 140060 }, { "epoch": 2.7990487990088324, "grad_norm": 0.11636366695165634, "learning_rate": 4.182736991600295e-06, "loss": 0.0179, "step": 140070 }, { "epoch": 2.799248631149834, "grad_norm": 0.17950452864170074, "learning_rate": 4.178600275535921e-06, "loss": 0.0291, "step": 140080 }, { "epoch": 2.799448463290836, "grad_norm": 0.11503332853317261, "learning_rate": 4.174465516877574e-06, "loss": 0.0325, "step": 140090 }, { "epoch": 2.799648295431837, "grad_norm": 0.11379482597112656, "learning_rate": 4.1703327158018724e-06, "loss": 0.019, "step": 140100 }, { "epoch": 2.799848127572839, "grad_norm": 0.09267362952232361, "learning_rate": 4.166201872485375e-06, "loss": 0.0272, "step": 140110 }, { "epoch": 2.80004795971384, "grad_norm": 0.1278940886259079, "learning_rate": 4.16207298710452e-06, "loss": 0.0242, "step": 140120 }, { "epoch": 2.800247791854842, "grad_norm": 0.09586586058139801, "learning_rate": 4.157946059835705e-06, "loss": 0.0158, "step": 140130 }, { "epoch": 2.8004476239958436, "grad_norm": 0.0875387042760849, "learning_rate": 4.153821090855231e-06, "loss": 0.0212, "step": 140140 }, { "epoch": 2.8006474561368453, "grad_norm": 0.1633542776107788, "learning_rate": 4.149698080339287e-06, "loss": 0.0228, "step": 140150 }, { "epoch": 2.8008472882778466, "grad_norm": 0.0835055261850357, "learning_rate": 4.145577028464015e-06, "loss": 0.0175, "step": 140160 }, { "epoch": 2.8010471204188483, "grad_norm": 0.09388340264558792, "learning_rate": 4.1414579354054514e-06, "loss": 0.0294, "step": 140170 }, { "epoch": 2.8012469525598496, "grad_norm": 0.7282178401947021, "learning_rate": 4.137340801339562e-06, "loss": 0.023, "step": 140180 }, { "epoch": 2.8014467847008513, "grad_norm": 0.1192418709397316, "learning_rate": 4.133225626442222e-06, "loss": 0.0238, "step": 140190 }, { "epoch": 2.801646616841853, "grad_norm": 0.10999360680580139, "learning_rate": 4.129112410889224e-06, "loss": 0.0282, "step": 140200 }, { "epoch": 2.8018464489828543, "grad_norm": 0.14859850704669952, "learning_rate": 4.1250011548562724e-06, "loss": 0.019, "step": 140210 }, { "epoch": 2.802046281123856, "grad_norm": 0.16241809725761414, "learning_rate": 4.120891858519e-06, "loss": 0.0243, "step": 140220 }, { "epoch": 2.8022461132648573, "grad_norm": 0.12430007755756378, "learning_rate": 4.116784522052941e-06, "loss": 0.0219, "step": 140230 }, { "epoch": 2.802445945405859, "grad_norm": 0.1055620014667511, "learning_rate": 4.112679145633569e-06, "loss": 0.0183, "step": 140240 }, { "epoch": 2.8026457775468607, "grad_norm": 0.11028183251619339, "learning_rate": 4.108575729436237e-06, "loss": 0.0325, "step": 140250 }, { "epoch": 2.8028456096878624, "grad_norm": 0.10931218415498734, "learning_rate": 4.104474273636233e-06, "loss": 0.0253, "step": 140260 }, { "epoch": 2.8030454418288637, "grad_norm": 0.13036836683750153, "learning_rate": 4.1003747784087946e-06, "loss": 0.0208, "step": 140270 }, { "epoch": 2.8032452739698654, "grad_norm": 0.12635256350040436, "learning_rate": 4.09627724392902e-06, "loss": 0.0246, "step": 140280 }, { "epoch": 2.8034451061108667, "grad_norm": 0.10086759924888611, "learning_rate": 4.092181670371948e-06, "loss": 0.0264, "step": 140290 }, { "epoch": 2.8036449382518684, "grad_norm": 0.21252621710300446, "learning_rate": 4.0880880579125425e-06, "loss": 0.0252, "step": 140300 }, { "epoch": 2.80384477039287, "grad_norm": 0.16680072247982025, "learning_rate": 4.083996406725671e-06, "loss": 0.0196, "step": 140310 }, { "epoch": 2.8040446025338714, "grad_norm": 0.12473295629024506, "learning_rate": 4.079906716986126e-06, "loss": 0.0257, "step": 140320 }, { "epoch": 2.804244434674873, "grad_norm": 0.12706921994686127, "learning_rate": 4.075818988868607e-06, "loss": 0.0205, "step": 140330 }, { "epoch": 2.8044442668158744, "grad_norm": 0.10289417952299118, "learning_rate": 4.071733222547741e-06, "loss": 0.0203, "step": 140340 }, { "epoch": 2.804644098956876, "grad_norm": 0.1144460141658783, "learning_rate": 4.067649418198044e-06, "loss": 0.0293, "step": 140350 }, { "epoch": 2.804843931097878, "grad_norm": 0.10795368999242783, "learning_rate": 4.0635675759939925e-06, "loss": 0.022, "step": 140360 }, { "epoch": 2.8050437632388796, "grad_norm": 0.13611768186092377, "learning_rate": 4.0594876961099534e-06, "loss": 0.0187, "step": 140370 }, { "epoch": 2.805243595379881, "grad_norm": 0.09389021247625351, "learning_rate": 4.055409778720193e-06, "loss": 0.0218, "step": 140380 }, { "epoch": 2.8054434275208826, "grad_norm": 0.11211718618869781, "learning_rate": 4.05133382399892e-06, "loss": 0.0276, "step": 140390 }, { "epoch": 2.805643259661884, "grad_norm": 0.11330904066562653, "learning_rate": 4.04725983212027e-06, "loss": 0.0274, "step": 140400 }, { "epoch": 2.8058430918028856, "grad_norm": 0.09713754057884216, "learning_rate": 4.043187803258258e-06, "loss": 0.0237, "step": 140410 }, { "epoch": 2.8060429239438873, "grad_norm": 0.12195655703544617, "learning_rate": 4.039117737586834e-06, "loss": 0.023, "step": 140420 }, { "epoch": 2.8062427560848886, "grad_norm": 0.1000877171754837, "learning_rate": 4.035049635279869e-06, "loss": 0.0244, "step": 140430 }, { "epoch": 2.8064425882258903, "grad_norm": 0.1344291865825653, "learning_rate": 4.030983496511142e-06, "loss": 0.0246, "step": 140440 }, { "epoch": 2.8066424203668916, "grad_norm": 0.16213653981685638, "learning_rate": 4.026919321454353e-06, "loss": 0.0282, "step": 140450 }, { "epoch": 2.8068422525078933, "grad_norm": 0.08543375134468079, "learning_rate": 4.022857110283113e-06, "loss": 0.0178, "step": 140460 }, { "epoch": 2.807042084648895, "grad_norm": 0.15181350708007812, "learning_rate": 4.018796863170965e-06, "loss": 0.0205, "step": 140470 }, { "epoch": 2.8072419167898968, "grad_norm": 0.14676594734191895, "learning_rate": 4.014738580291327e-06, "loss": 0.0248, "step": 140480 }, { "epoch": 2.807441748930898, "grad_norm": 0.0969889834523201, "learning_rate": 4.010682261817589e-06, "loss": 0.016, "step": 140490 }, { "epoch": 2.8076415810718998, "grad_norm": 0.14875799417495728, "learning_rate": 4.0066279079230216e-06, "loss": 0.021, "step": 140500 }, { "epoch": 2.807841413212901, "grad_norm": 0.10977966338396072, "learning_rate": 4.002575518780821e-06, "loss": 0.0233, "step": 140510 }, { "epoch": 2.8080412453539028, "grad_norm": 0.126816987991333, "learning_rate": 3.998525094564087e-06, "loss": 0.0233, "step": 140520 }, { "epoch": 2.8082410774949045, "grad_norm": 0.11325331032276154, "learning_rate": 3.994476635445848e-06, "loss": 0.0162, "step": 140530 }, { "epoch": 2.8084409096359058, "grad_norm": 0.16809405386447906, "learning_rate": 3.9904301415990665e-06, "loss": 0.0247, "step": 140540 }, { "epoch": 2.8086407417769075, "grad_norm": 0.12367209792137146, "learning_rate": 3.98638561319658e-06, "loss": 0.0185, "step": 140550 }, { "epoch": 2.8088405739179088, "grad_norm": 0.11756930500268936, "learning_rate": 3.982343050411169e-06, "loss": 0.0276, "step": 140560 }, { "epoch": 2.8090404060589105, "grad_norm": 0.14568407833576202, "learning_rate": 3.978302453415528e-06, "loss": 0.0232, "step": 140570 }, { "epoch": 2.809240238199912, "grad_norm": 0.10036627948284149, "learning_rate": 3.974263822382262e-06, "loss": 0.024, "step": 140580 }, { "epoch": 2.8094400703409135, "grad_norm": 0.08110202103853226, "learning_rate": 3.970227157483891e-06, "loss": 0.0208, "step": 140590 }, { "epoch": 2.809639902481915, "grad_norm": 0.09336075931787491, "learning_rate": 3.966192458892859e-06, "loss": 0.0196, "step": 140600 }, { "epoch": 2.8098397346229165, "grad_norm": 0.14294478297233582, "learning_rate": 3.962159726781522e-06, "loss": 0.0253, "step": 140610 }, { "epoch": 2.810039566763918, "grad_norm": 0.12149154394865036, "learning_rate": 3.958128961322138e-06, "loss": 0.0265, "step": 140620 }, { "epoch": 2.81023939890492, "grad_norm": 0.09946602582931519, "learning_rate": 3.954100162686908e-06, "loss": 0.022, "step": 140630 }, { "epoch": 2.8104392310459216, "grad_norm": 0.1417643427848816, "learning_rate": 3.9500733310479374e-06, "loss": 0.0242, "step": 140640 }, { "epoch": 2.810639063186923, "grad_norm": 0.1144004762172699, "learning_rate": 3.946048466577229e-06, "loss": 0.0206, "step": 140650 }, { "epoch": 2.8108388953279246, "grad_norm": 0.12795992195606232, "learning_rate": 3.942025569446717e-06, "loss": 0.0149, "step": 140660 }, { "epoch": 2.811038727468926, "grad_norm": 0.15143491327762604, "learning_rate": 3.93800463982828e-06, "loss": 0.0209, "step": 140670 }, { "epoch": 2.8112385596099276, "grad_norm": 0.19296537339687347, "learning_rate": 3.933985677893659e-06, "loss": 0.0218, "step": 140680 }, { "epoch": 2.8114383917509294, "grad_norm": 0.09194588661193848, "learning_rate": 3.929968683814539e-06, "loss": 0.0229, "step": 140690 }, { "epoch": 2.8116382238919306, "grad_norm": 0.11831450462341309, "learning_rate": 3.925953657762526e-06, "loss": 0.0272, "step": 140700 }, { "epoch": 2.8118380560329324, "grad_norm": 0.1277034431695938, "learning_rate": 3.921940599909135e-06, "loss": 0.0194, "step": 140710 }, { "epoch": 2.8120378881739336, "grad_norm": 0.10663869231939316, "learning_rate": 3.917929510425794e-06, "loss": 0.0252, "step": 140720 }, { "epoch": 2.8122377203149354, "grad_norm": 0.13239289820194244, "learning_rate": 3.913920389483844e-06, "loss": 0.0216, "step": 140730 }, { "epoch": 2.812437552455937, "grad_norm": 0.12100798636674881, "learning_rate": 3.909913237254564e-06, "loss": 0.0175, "step": 140740 }, { "epoch": 2.812637384596939, "grad_norm": 0.12317465245723724, "learning_rate": 3.905908053909102e-06, "loss": 0.0252, "step": 140750 }, { "epoch": 2.81283721673794, "grad_norm": 0.0928434506058693, "learning_rate": 3.901904839618581e-06, "loss": 0.0195, "step": 140760 }, { "epoch": 2.813037048878942, "grad_norm": 0.13057492673397064, "learning_rate": 3.897903594554003e-06, "loss": 0.0192, "step": 140770 }, { "epoch": 2.813236881019943, "grad_norm": 0.1428440511226654, "learning_rate": 3.893904318886288e-06, "loss": 0.0237, "step": 140780 }, { "epoch": 2.813436713160945, "grad_norm": 0.13664275407791138, "learning_rate": 3.8899070127862815e-06, "loss": 0.0194, "step": 140790 }, { "epoch": 2.8136365453019465, "grad_norm": 0.09406275302171707, "learning_rate": 3.885911676424736e-06, "loss": 0.0235, "step": 140800 }, { "epoch": 2.813836377442948, "grad_norm": 0.11173649877309799, "learning_rate": 3.881918309972332e-06, "loss": 0.0336, "step": 140810 }, { "epoch": 2.8140362095839495, "grad_norm": 0.11914002895355225, "learning_rate": 3.8779269135996556e-06, "loss": 0.0242, "step": 140820 }, { "epoch": 2.814236041724951, "grad_norm": 0.14792920649051666, "learning_rate": 3.873937487477214e-06, "loss": 0.021, "step": 140830 }, { "epoch": 2.8144358738659525, "grad_norm": 0.10265066474676132, "learning_rate": 3.869950031775426e-06, "loss": 0.033, "step": 140840 }, { "epoch": 2.8146357060069542, "grad_norm": 0.19961218535900116, "learning_rate": 3.865964546664624e-06, "loss": 0.0208, "step": 140850 }, { "epoch": 2.814835538147956, "grad_norm": 0.07223838567733765, "learning_rate": 3.861981032315071e-06, "loss": 0.0288, "step": 140860 }, { "epoch": 2.8150353702889572, "grad_norm": 0.1099616289138794, "learning_rate": 3.857999488896941e-06, "loss": 0.021, "step": 140870 }, { "epoch": 2.815235202429959, "grad_norm": 0.09895453602075577, "learning_rate": 3.854019916580293e-06, "loss": 0.0226, "step": 140880 }, { "epoch": 2.8154350345709602, "grad_norm": 0.11650446057319641, "learning_rate": 3.850042315535135e-06, "loss": 0.0256, "step": 140890 }, { "epoch": 2.815634866711962, "grad_norm": 0.11230450868606567, "learning_rate": 3.846066685931399e-06, "loss": 0.0234, "step": 140900 }, { "epoch": 2.8158346988529637, "grad_norm": 0.13938578963279724, "learning_rate": 3.842093027938914e-06, "loss": 0.0265, "step": 140910 }, { "epoch": 2.816034530993965, "grad_norm": 0.13765159249305725, "learning_rate": 3.838121341727413e-06, "loss": 0.028, "step": 140920 }, { "epoch": 2.8162343631349667, "grad_norm": 0.06862598657608032, "learning_rate": 3.834151627466554e-06, "loss": 0.0302, "step": 140930 }, { "epoch": 2.816434195275968, "grad_norm": 0.11150835454463959, "learning_rate": 3.830183885325949e-06, "loss": 0.0338, "step": 140940 }, { "epoch": 2.8166340274169697, "grad_norm": 0.12188371270895004, "learning_rate": 3.826218115475061e-06, "loss": 0.0198, "step": 140950 }, { "epoch": 2.8168338595579714, "grad_norm": 0.14380720257759094, "learning_rate": 3.822254318083313e-06, "loss": 0.0243, "step": 140960 }, { "epoch": 2.817033691698973, "grad_norm": 0.15392613410949707, "learning_rate": 3.81829249332003e-06, "loss": 0.0234, "step": 140970 }, { "epoch": 2.8172335238399744, "grad_norm": 0.1346862018108368, "learning_rate": 3.8143326413544513e-06, "loss": 0.0181, "step": 140980 }, { "epoch": 2.817433355980976, "grad_norm": 0.11364427208900452, "learning_rate": 3.810374762355734e-06, "loss": 0.0211, "step": 140990 }, { "epoch": 2.8176331881219774, "grad_norm": 0.10464683920145035, "learning_rate": 3.806418856492955e-06, "loss": 0.0187, "step": 141000 }, { "epoch": 2.817833020262979, "grad_norm": 0.12397512793540955, "learning_rate": 3.8024649239351116e-06, "loss": 0.0175, "step": 141010 }, { "epoch": 2.818032852403981, "grad_norm": 0.09655497223138809, "learning_rate": 3.7985129648510866e-06, "loss": 0.0182, "step": 141020 }, { "epoch": 2.818232684544982, "grad_norm": 0.08560671657323837, "learning_rate": 3.794562979409716e-06, "loss": 0.0287, "step": 141030 }, { "epoch": 2.818432516685984, "grad_norm": 0.1371065378189087, "learning_rate": 3.7906149677797433e-06, "loss": 0.0213, "step": 141040 }, { "epoch": 2.818632348826985, "grad_norm": 0.11088087409734726, "learning_rate": 3.7866689301297996e-06, "loss": 0.0191, "step": 141050 }, { "epoch": 2.818832180967987, "grad_norm": 0.1326799988746643, "learning_rate": 3.782724866628451e-06, "loss": 0.0207, "step": 141060 }, { "epoch": 2.8190320131089885, "grad_norm": 0.08728016912937164, "learning_rate": 3.7787827774442174e-06, "loss": 0.0238, "step": 141070 }, { "epoch": 2.8192318452499903, "grad_norm": 0.0851711705327034, "learning_rate": 3.774842662745459e-06, "loss": 0.0188, "step": 141080 }, { "epoch": 2.8194316773909915, "grad_norm": 0.14661771059036255, "learning_rate": 3.7709045227005025e-06, "loss": 0.0284, "step": 141090 }, { "epoch": 2.8196315095319933, "grad_norm": 0.08409136533737183, "learning_rate": 3.766968357477579e-06, "loss": 0.034, "step": 141100 }, { "epoch": 2.8198313416729945, "grad_norm": 0.11274319142103195, "learning_rate": 3.763034167244839e-06, "loss": 0.025, "step": 141110 }, { "epoch": 2.8200311738139963, "grad_norm": 0.11655137687921524, "learning_rate": 3.7591019521703354e-06, "loss": 0.0246, "step": 141120 }, { "epoch": 2.820231005954998, "grad_norm": 0.09561891853809357, "learning_rate": 3.755171712422051e-06, "loss": 0.0212, "step": 141130 }, { "epoch": 2.8204308380959993, "grad_norm": 0.16420789062976837, "learning_rate": 3.7512434481678847e-06, "loss": 0.0308, "step": 141140 }, { "epoch": 2.820630670237001, "grad_norm": 0.1059422567486763, "learning_rate": 3.7473171595756196e-06, "loss": 0.0241, "step": 141150 }, { "epoch": 2.8208305023780023, "grad_norm": 0.1182713508605957, "learning_rate": 3.74339284681301e-06, "loss": 0.0232, "step": 141160 }, { "epoch": 2.821030334519004, "grad_norm": 0.09890193492174149, "learning_rate": 3.7394705100476824e-06, "loss": 0.028, "step": 141170 }, { "epoch": 2.8212301666600057, "grad_norm": 0.19441023468971252, "learning_rate": 3.7355501494471866e-06, "loss": 0.028, "step": 141180 }, { "epoch": 2.8214299988010074, "grad_norm": 0.14292073249816895, "learning_rate": 3.731631765179e-06, "loss": 0.0203, "step": 141190 }, { "epoch": 2.8216298309420087, "grad_norm": 0.09933631122112274, "learning_rate": 3.7277153574104993e-06, "loss": 0.0271, "step": 141200 }, { "epoch": 2.8218296630830104, "grad_norm": 0.0890156701207161, "learning_rate": 3.7238009263090014e-06, "loss": 0.0209, "step": 141210 }, { "epoch": 2.8220294952240117, "grad_norm": 0.15091422200202942, "learning_rate": 3.7198884720417103e-06, "loss": 0.0255, "step": 141220 }, { "epoch": 2.8222293273650134, "grad_norm": 0.10962489992380142, "learning_rate": 3.7159779947757713e-06, "loss": 0.0193, "step": 141230 }, { "epoch": 2.822429159506015, "grad_norm": 0.1570911705493927, "learning_rate": 3.7120694946782164e-06, "loss": 0.0224, "step": 141240 }, { "epoch": 2.8226289916470164, "grad_norm": 0.11303620040416718, "learning_rate": 3.7081629719160237e-06, "loss": 0.0244, "step": 141250 }, { "epoch": 2.822828823788018, "grad_norm": 0.11226882040500641, "learning_rate": 3.704258426656071e-06, "loss": 0.0264, "step": 141260 }, { "epoch": 2.8230286559290194, "grad_norm": 0.11661660671234131, "learning_rate": 3.700355859065158e-06, "loss": 0.028, "step": 141270 }, { "epoch": 2.823228488070021, "grad_norm": 0.10487830638885498, "learning_rate": 3.696455269309973e-06, "loss": 0.0206, "step": 141280 }, { "epoch": 2.823428320211023, "grad_norm": 0.10540905594825745, "learning_rate": 3.6925566575571556e-06, "loss": 0.029, "step": 141290 }, { "epoch": 2.823628152352024, "grad_norm": 0.11840128153562546, "learning_rate": 3.6886600239732505e-06, "loss": 0.0211, "step": 141300 }, { "epoch": 2.823827984493026, "grad_norm": 0.1109454482793808, "learning_rate": 3.684765368724724e-06, "loss": 0.0263, "step": 141310 }, { "epoch": 2.8240278166340276, "grad_norm": 0.15921184420585632, "learning_rate": 3.6808726919779325e-06, "loss": 0.0304, "step": 141320 }, { "epoch": 2.824227648775029, "grad_norm": 0.13392281532287598, "learning_rate": 3.6769819938991533e-06, "loss": 0.0258, "step": 141330 }, { "epoch": 2.8244274809160306, "grad_norm": 0.15069012343883514, "learning_rate": 3.6730932746546264e-06, "loss": 0.0306, "step": 141340 }, { "epoch": 2.8246273130570323, "grad_norm": 0.09499844163656235, "learning_rate": 3.669206534410441e-06, "loss": 0.0241, "step": 141350 }, { "epoch": 2.8248271451980336, "grad_norm": 0.11882045120000839, "learning_rate": 3.665321773332636e-06, "loss": 0.0168, "step": 141360 }, { "epoch": 2.8250269773390353, "grad_norm": 0.13348299264907837, "learning_rate": 3.661438991587168e-06, "loss": 0.0236, "step": 141370 }, { "epoch": 2.8252268094800366, "grad_norm": 0.13265067338943481, "learning_rate": 3.6575581893399048e-06, "loss": 0.0254, "step": 141380 }, { "epoch": 2.8254266416210383, "grad_norm": 0.13167576491832733, "learning_rate": 3.6536793667566128e-06, "loss": 0.0247, "step": 141390 }, { "epoch": 2.82562647376204, "grad_norm": 0.10928138345479965, "learning_rate": 3.6498025240030045e-06, "loss": 0.0248, "step": 141400 }, { "epoch": 2.8258263059030413, "grad_norm": 0.09442759305238724, "learning_rate": 3.6459276612446913e-06, "loss": 0.0249, "step": 141410 }, { "epoch": 2.826026138044043, "grad_norm": 0.15411055088043213, "learning_rate": 3.642054778647175e-06, "loss": 0.0289, "step": 141420 }, { "epoch": 2.8262259701850443, "grad_norm": 0.12093519419431686, "learning_rate": 3.6381838763759222e-06, "loss": 0.0304, "step": 141430 }, { "epoch": 2.826425802326046, "grad_norm": 0.152613565325737, "learning_rate": 3.6343149545962963e-06, "loss": 0.0221, "step": 141440 }, { "epoch": 2.8266256344670477, "grad_norm": 0.17258320748806, "learning_rate": 3.6304480134735473e-06, "loss": 0.0342, "step": 141450 }, { "epoch": 2.8268254666080495, "grad_norm": 0.112763911485672, "learning_rate": 3.6265830531728828e-06, "loss": 0.0282, "step": 141460 }, { "epoch": 2.8270252987490507, "grad_norm": 0.15262548625469208, "learning_rate": 3.6227200738593924e-06, "loss": 0.0182, "step": 141470 }, { "epoch": 2.8272251308900525, "grad_norm": 0.09656722843647003, "learning_rate": 3.6188590756981056e-06, "loss": 0.0265, "step": 141480 }, { "epoch": 2.8274249630310537, "grad_norm": 0.11780364066362381, "learning_rate": 3.6150000588539567e-06, "loss": 0.0204, "step": 141490 }, { "epoch": 2.8276247951720554, "grad_norm": 0.11571335792541504, "learning_rate": 3.6111430234917864e-06, "loss": 0.0234, "step": 141500 }, { "epoch": 2.827824627313057, "grad_norm": 0.13714320957660675, "learning_rate": 3.607287969776374e-06, "loss": 0.0186, "step": 141510 }, { "epoch": 2.8280244594540584, "grad_norm": 0.1005249097943306, "learning_rate": 3.6034348978723876e-06, "loss": 0.0309, "step": 141520 }, { "epoch": 2.82822429159506, "grad_norm": 0.09242361038923264, "learning_rate": 3.5995838079444346e-06, "loss": 0.0201, "step": 141530 }, { "epoch": 2.8284241237360614, "grad_norm": 0.0997653529047966, "learning_rate": 3.5957347001570274e-06, "loss": 0.0236, "step": 141540 }, { "epoch": 2.828623955877063, "grad_norm": 0.1620521992444992, "learning_rate": 3.5918875746745795e-06, "loss": 0.0246, "step": 141550 }, { "epoch": 2.828823788018065, "grad_norm": 0.15961141884326935, "learning_rate": 3.5880424316614257e-06, "loss": 0.0217, "step": 141560 }, { "epoch": 2.8290236201590666, "grad_norm": 0.12059716880321503, "learning_rate": 3.584199271281863e-06, "loss": 0.0269, "step": 141570 }, { "epoch": 2.829223452300068, "grad_norm": 0.1324651539325714, "learning_rate": 3.5803580937000257e-06, "loss": 0.0235, "step": 141580 }, { "epoch": 2.8294232844410696, "grad_norm": 0.11412351578474045, "learning_rate": 3.576518899080011e-06, "loss": 0.0254, "step": 141590 }, { "epoch": 2.829623116582071, "grad_norm": 0.09233610332012177, "learning_rate": 3.572681687585833e-06, "loss": 0.0211, "step": 141600 }, { "epoch": 2.8298229487230726, "grad_norm": 0.11976239085197449, "learning_rate": 3.568846459381403e-06, "loss": 0.0325, "step": 141610 }, { "epoch": 2.8300227808640743, "grad_norm": 0.12147387117147446, "learning_rate": 3.5650132146305525e-06, "loss": 0.0337, "step": 141620 }, { "epoch": 2.8302226130050756, "grad_norm": 0.10171720385551453, "learning_rate": 3.561181953497034e-06, "loss": 0.0214, "step": 141630 }, { "epoch": 2.8304224451460773, "grad_norm": 0.11596839129924774, "learning_rate": 3.5573526761445206e-06, "loss": 0.0259, "step": 141640 }, { "epoch": 2.8306222772870786, "grad_norm": 0.13360781967639923, "learning_rate": 3.5535253827365666e-06, "loss": 0.03, "step": 141650 }, { "epoch": 2.8308221094280803, "grad_norm": 0.12203604727983475, "learning_rate": 3.5497000734366957e-06, "loss": 0.0217, "step": 141660 }, { "epoch": 2.831021941569082, "grad_norm": 0.10159782320261002, "learning_rate": 3.5458767484083056e-06, "loss": 0.0207, "step": 141670 }, { "epoch": 2.8312217737100838, "grad_norm": 0.1521148830652237, "learning_rate": 3.5420554078147204e-06, "loss": 0.0312, "step": 141680 }, { "epoch": 2.831421605851085, "grad_norm": 0.12144622206687927, "learning_rate": 3.5382360518191714e-06, "loss": 0.0241, "step": 141690 }, { "epoch": 2.8316214379920868, "grad_norm": 0.10208268463611603, "learning_rate": 3.5344186805848335e-06, "loss": 0.0337, "step": 141700 }, { "epoch": 2.831821270133088, "grad_norm": 0.10225149244070053, "learning_rate": 3.530603294274776e-06, "loss": 0.0172, "step": 141710 }, { "epoch": 2.8320211022740898, "grad_norm": 0.11480185389518738, "learning_rate": 3.526789893051974e-06, "loss": 0.0276, "step": 141720 }, { "epoch": 2.8322209344150915, "grad_norm": 0.1465948522090912, "learning_rate": 3.5229784770793307e-06, "loss": 0.0258, "step": 141730 }, { "epoch": 2.8324207665560928, "grad_norm": 0.1086023673415184, "learning_rate": 3.5191690465196715e-06, "loss": 0.021, "step": 141740 }, { "epoch": 2.8326205986970945, "grad_norm": 0.14103160798549652, "learning_rate": 3.515361601535716e-06, "loss": 0.0254, "step": 141750 }, { "epoch": 2.8328204308380958, "grad_norm": 0.11982002854347229, "learning_rate": 3.5115561422901235e-06, "loss": 0.029, "step": 141760 }, { "epoch": 2.8330202629790975, "grad_norm": 0.19166415929794312, "learning_rate": 3.5077526689454467e-06, "loss": 0.0218, "step": 141770 }, { "epoch": 2.833220095120099, "grad_norm": 0.09665720909833908, "learning_rate": 3.5039511816641667e-06, "loss": 0.0184, "step": 141780 }, { "epoch": 2.833419927261101, "grad_norm": 0.13704504072666168, "learning_rate": 3.5001516806086766e-06, "loss": 0.0248, "step": 141790 }, { "epoch": 2.833619759402102, "grad_norm": 0.1366010457277298, "learning_rate": 3.4963541659412846e-06, "loss": 0.021, "step": 141800 }, { "epoch": 2.833819591543104, "grad_norm": 0.14944525063037872, "learning_rate": 3.4925586378242224e-06, "loss": 0.0284, "step": 141810 }, { "epoch": 2.834019423684105, "grad_norm": 0.12437988072633743, "learning_rate": 3.488765096419605e-06, "loss": 0.0227, "step": 141820 }, { "epoch": 2.834219255825107, "grad_norm": 0.1343207210302353, "learning_rate": 3.484973541889497e-06, "loss": 0.0191, "step": 141830 }, { "epoch": 2.8344190879661086, "grad_norm": 0.13709378242492676, "learning_rate": 3.48118397439588e-06, "loss": 0.0243, "step": 141840 }, { "epoch": 2.83461892010711, "grad_norm": 0.09730903059244156, "learning_rate": 3.4773963941006184e-06, "loss": 0.0263, "step": 141850 }, { "epoch": 2.8348187522481116, "grad_norm": 0.10996086895465851, "learning_rate": 3.4736108011655223e-06, "loss": 0.025, "step": 141860 }, { "epoch": 2.835018584389113, "grad_norm": 0.1267790049314499, "learning_rate": 3.4698271957522956e-06, "loss": 0.0315, "step": 141870 }, { "epoch": 2.8352184165301146, "grad_norm": 0.13219621777534485, "learning_rate": 3.4660455780225754e-06, "loss": 0.0217, "step": 141880 }, { "epoch": 2.8354182486711164, "grad_norm": 0.09279389679431915, "learning_rate": 3.462265948137905e-06, "loss": 0.017, "step": 141890 }, { "epoch": 2.835618080812118, "grad_norm": 0.10124363750219345, "learning_rate": 3.4584883062597386e-06, "loss": 0.0167, "step": 141900 }, { "epoch": 2.8358179129531194, "grad_norm": 0.15097633004188538, "learning_rate": 3.454712652549458e-06, "loss": 0.0201, "step": 141910 }, { "epoch": 2.836017745094121, "grad_norm": 0.12714889645576477, "learning_rate": 3.4509389871683283e-06, "loss": 0.0226, "step": 141920 }, { "epoch": 2.8362175772351224, "grad_norm": 0.0701952576637268, "learning_rate": 3.447167310277588e-06, "loss": 0.0221, "step": 141930 }, { "epoch": 2.836417409376124, "grad_norm": 0.1220274344086647, "learning_rate": 3.4433976220383414e-06, "loss": 0.0275, "step": 141940 }, { "epoch": 2.836617241517126, "grad_norm": 0.172810897231102, "learning_rate": 3.4396299226116145e-06, "loss": 0.0246, "step": 141950 }, { "epoch": 2.836817073658127, "grad_norm": 0.09482385963201523, "learning_rate": 3.4358642121583572e-06, "loss": 0.0255, "step": 141960 }, { "epoch": 2.837016905799129, "grad_norm": 0.10835149139165878, "learning_rate": 3.4321004908394516e-06, "loss": 0.0236, "step": 141970 }, { "epoch": 2.83721673794013, "grad_norm": 0.11812600493431091, "learning_rate": 3.428338758815658e-06, "loss": 0.0189, "step": 141980 }, { "epoch": 2.837416570081132, "grad_norm": 0.14786460995674133, "learning_rate": 3.424579016247681e-06, "loss": 0.0217, "step": 141990 }, { "epoch": 2.8376164022221335, "grad_norm": 0.1175513043999672, "learning_rate": 3.4208212632961258e-06, "loss": 0.0279, "step": 142000 }, { "epoch": 2.8378162343631352, "grad_norm": 0.0676087960600853, "learning_rate": 3.4170655001215192e-06, "loss": 0.0151, "step": 142010 }, { "epoch": 2.8380160665041365, "grad_norm": 0.1312078982591629, "learning_rate": 3.413311726884294e-06, "loss": 0.0242, "step": 142020 }, { "epoch": 2.8382158986451382, "grad_norm": 0.051622163504362106, "learning_rate": 3.4095599437448113e-06, "loss": 0.018, "step": 142030 }, { "epoch": 2.8384157307861395, "grad_norm": 0.1521049439907074, "learning_rate": 3.405810150863348e-06, "loss": 0.0258, "step": 142040 }, { "epoch": 2.8386155629271412, "grad_norm": 0.10428927838802338, "learning_rate": 3.402062348400059e-06, "loss": 0.0268, "step": 142050 }, { "epoch": 2.838815395068143, "grad_norm": 0.18342134356498718, "learning_rate": 3.3983165365150727e-06, "loss": 0.0229, "step": 142060 }, { "epoch": 2.8390152272091442, "grad_norm": 0.12806235253810883, "learning_rate": 3.394572715368405e-06, "loss": 0.0318, "step": 142070 }, { "epoch": 2.839215059350146, "grad_norm": 0.1523423194885254, "learning_rate": 3.3908308851199665e-06, "loss": 0.034, "step": 142080 }, { "epoch": 2.8394148914911472, "grad_norm": 0.15995196998119354, "learning_rate": 3.387091045929597e-06, "loss": 0.026, "step": 142090 }, { "epoch": 2.839614723632149, "grad_norm": 0.09928761422634125, "learning_rate": 3.3833531979570786e-06, "loss": 0.0248, "step": 142100 }, { "epoch": 2.8398145557731507, "grad_norm": 0.10745925456285477, "learning_rate": 3.379617341362079e-06, "loss": 0.0249, "step": 142110 }, { "epoch": 2.840014387914152, "grad_norm": 0.12867505848407745, "learning_rate": 3.375883476304176e-06, "loss": 0.0229, "step": 142120 }, { "epoch": 2.8402142200551537, "grad_norm": 0.11879383027553558, "learning_rate": 3.3721516029428803e-06, "loss": 0.0257, "step": 142130 }, { "epoch": 2.8404140521961554, "grad_norm": 0.1606573909521103, "learning_rate": 3.368421721437609e-06, "loss": 0.0155, "step": 142140 }, { "epoch": 2.8406138843371567, "grad_norm": 0.13866275548934937, "learning_rate": 3.3646938319476964e-06, "loss": 0.0173, "step": 142150 }, { "epoch": 2.8408137164781584, "grad_norm": 0.09718848019838333, "learning_rate": 3.360967934632392e-06, "loss": 0.0351, "step": 142160 }, { "epoch": 2.84101354861916, "grad_norm": 0.11651401966810226, "learning_rate": 3.3572440296508577e-06, "loss": 0.0239, "step": 142170 }, { "epoch": 2.8412133807601614, "grad_norm": 0.11748199909925461, "learning_rate": 3.3535221171621723e-06, "loss": 0.0238, "step": 142180 }, { "epoch": 2.841413212901163, "grad_norm": 0.07697159051895142, "learning_rate": 3.34980219732533e-06, "loss": 0.0232, "step": 142190 }, { "epoch": 2.8416130450421644, "grad_norm": 0.10402780026197433, "learning_rate": 3.3460842702992435e-06, "loss": 0.0244, "step": 142200 }, { "epoch": 2.841812877183166, "grad_norm": 0.22822725772857666, "learning_rate": 3.342368336242735e-06, "loss": 0.0217, "step": 142210 }, { "epoch": 2.842012709324168, "grad_norm": 0.11515304446220398, "learning_rate": 3.3386543953145276e-06, "loss": 0.0312, "step": 142220 }, { "epoch": 2.842212541465169, "grad_norm": 0.09890434890985489, "learning_rate": 3.3349424476732836e-06, "loss": 0.0307, "step": 142230 }, { "epoch": 2.842412373606171, "grad_norm": 0.1395050287246704, "learning_rate": 3.3312324934775817e-06, "loss": 0.0227, "step": 142240 }, { "epoch": 2.842612205747172, "grad_norm": 0.11847654730081558, "learning_rate": 3.3275245328858897e-06, "loss": 0.0252, "step": 142250 }, { "epoch": 2.842812037888174, "grad_norm": 0.11086561530828476, "learning_rate": 3.323818566056608e-06, "loss": 0.0204, "step": 142260 }, { "epoch": 2.8430118700291755, "grad_norm": 0.10558376461267471, "learning_rate": 3.320114593148049e-06, "loss": 0.0169, "step": 142270 }, { "epoch": 2.8432117021701773, "grad_norm": 0.13561588525772095, "learning_rate": 3.3164126143184483e-06, "loss": 0.0227, "step": 142280 }, { "epoch": 2.8434115343111785, "grad_norm": 0.12490549683570862, "learning_rate": 3.3127126297259335e-06, "loss": 0.0289, "step": 142290 }, { "epoch": 2.8436113664521803, "grad_norm": 0.09000261127948761, "learning_rate": 3.3090146395285683e-06, "loss": 0.0204, "step": 142300 }, { "epoch": 2.8438111985931815, "grad_norm": 0.09574836492538452, "learning_rate": 3.3053186438843307e-06, "loss": 0.0234, "step": 142310 }, { "epoch": 2.8440110307341833, "grad_norm": 0.11163614690303802, "learning_rate": 3.3016246429510898e-06, "loss": 0.0289, "step": 142320 }, { "epoch": 2.844210862875185, "grad_norm": 0.10602445900440216, "learning_rate": 3.297932636886658e-06, "loss": 0.0243, "step": 142330 }, { "epoch": 2.8444106950161863, "grad_norm": 0.12527023255825043, "learning_rate": 3.2942426258487593e-06, "loss": 0.0279, "step": 142340 }, { "epoch": 2.844610527157188, "grad_norm": 0.10204111784696579, "learning_rate": 3.2905546099950057e-06, "loss": 0.0212, "step": 142350 }, { "epoch": 2.8448103592981893, "grad_norm": 0.12574657797813416, "learning_rate": 3.2868685894829444e-06, "loss": 0.0221, "step": 142360 }, { "epoch": 2.845010191439191, "grad_norm": 0.13228043913841248, "learning_rate": 3.2831845644700598e-06, "loss": 0.0265, "step": 142370 }, { "epoch": 2.8452100235801927, "grad_norm": 0.14113256335258484, "learning_rate": 3.2795025351136986e-06, "loss": 0.0309, "step": 142380 }, { "epoch": 2.8454098557211944, "grad_norm": 0.09970297664403915, "learning_rate": 3.275822501571163e-06, "loss": 0.0197, "step": 142390 }, { "epoch": 2.8456096878621957, "grad_norm": 0.12980452179908752, "learning_rate": 3.2721444639996547e-06, "loss": 0.021, "step": 142400 }, { "epoch": 2.8458095200031974, "grad_norm": 0.11624373495578766, "learning_rate": 3.2684684225562924e-06, "loss": 0.0205, "step": 142410 }, { "epoch": 2.8460093521441987, "grad_norm": 0.09093902260065079, "learning_rate": 3.2647943773981117e-06, "loss": 0.0243, "step": 142420 }, { "epoch": 2.8462091842852004, "grad_norm": 0.10371606796979904, "learning_rate": 3.2611223286820535e-06, "loss": 0.0329, "step": 142430 }, { "epoch": 2.846409016426202, "grad_norm": 0.10034901648759842, "learning_rate": 3.257452276565004e-06, "loss": 0.0267, "step": 142440 }, { "epoch": 2.8466088485672034, "grad_norm": 0.201881542801857, "learning_rate": 3.253784221203704e-06, "loss": 0.0303, "step": 142450 }, { "epoch": 2.846808680708205, "grad_norm": 0.12960439920425415, "learning_rate": 3.250118162754873e-06, "loss": 0.019, "step": 142460 }, { "epoch": 2.8470085128492064, "grad_norm": 0.11085516959428787, "learning_rate": 3.2464541013751183e-06, "loss": 0.0198, "step": 142470 }, { "epoch": 2.847208344990208, "grad_norm": 0.10609413683414459, "learning_rate": 3.242792037220954e-06, "loss": 0.0247, "step": 142480 }, { "epoch": 2.84740817713121, "grad_norm": 0.0913047268986702, "learning_rate": 3.2391319704488167e-06, "loss": 0.0196, "step": 142490 }, { "epoch": 2.8476080092722116, "grad_norm": 0.14766620099544525, "learning_rate": 3.2354739012150526e-06, "loss": 0.0217, "step": 142500 }, { "epoch": 2.847807841413213, "grad_norm": 0.13111621141433716, "learning_rate": 3.2318178296759483e-06, "loss": 0.0297, "step": 142510 }, { "epoch": 2.8480076735542146, "grad_norm": 0.16109856963157654, "learning_rate": 3.2281637559876622e-06, "loss": 0.021, "step": 142520 }, { "epoch": 2.848207505695216, "grad_norm": 0.15990902483463287, "learning_rate": 3.224511680306297e-06, "loss": 0.0347, "step": 142530 }, { "epoch": 2.8484073378362176, "grad_norm": 0.07817641645669937, "learning_rate": 3.220861602787861e-06, "loss": 0.0208, "step": 142540 }, { "epoch": 2.8486071699772193, "grad_norm": 0.10759031027555466, "learning_rate": 3.217213523588286e-06, "loss": 0.0214, "step": 142550 }, { "epoch": 2.8488070021182206, "grad_norm": 0.12311133742332458, "learning_rate": 3.213567442863402e-06, "loss": 0.0218, "step": 142560 }, { "epoch": 2.8490068342592223, "grad_norm": 0.1265767514705658, "learning_rate": 3.2099233607689683e-06, "loss": 0.0202, "step": 142570 }, { "epoch": 2.8492066664002236, "grad_norm": 0.11242341995239258, "learning_rate": 3.2062812774606655e-06, "loss": 0.0273, "step": 142580 }, { "epoch": 2.8494064985412253, "grad_norm": 0.07725805789232254, "learning_rate": 3.2026411930940413e-06, "loss": 0.0217, "step": 142590 }, { "epoch": 2.849606330682227, "grad_norm": 0.1285472959280014, "learning_rate": 3.1990031078246274e-06, "loss": 0.0331, "step": 142600 }, { "epoch": 2.8498061628232287, "grad_norm": 0.12345512956380844, "learning_rate": 3.1953670218078323e-06, "loss": 0.0201, "step": 142610 }, { "epoch": 2.85000599496423, "grad_norm": 0.13788962364196777, "learning_rate": 3.1917329351989655e-06, "loss": 0.0265, "step": 142620 }, { "epoch": 2.8502058271052317, "grad_norm": 0.11141230165958405, "learning_rate": 3.188100848153275e-06, "loss": 0.0335, "step": 142630 }, { "epoch": 2.850405659246233, "grad_norm": 0.23593564331531525, "learning_rate": 3.1844707608259304e-06, "loss": 0.0227, "step": 142640 }, { "epoch": 2.8506054913872347, "grad_norm": 0.21687749028205872, "learning_rate": 3.1808426733719866e-06, "loss": 0.0237, "step": 142650 }, { "epoch": 2.8508053235282365, "grad_norm": 0.12938426434993744, "learning_rate": 3.1772165859464354e-06, "loss": 0.0278, "step": 142660 }, { "epoch": 2.8510051556692377, "grad_norm": 0.19300363957881927, "learning_rate": 3.1735924987041754e-06, "loss": 0.0234, "step": 142670 }, { "epoch": 2.8512049878102395, "grad_norm": 0.09109707176685333, "learning_rate": 3.1699704118000163e-06, "loss": 0.0201, "step": 142680 }, { "epoch": 2.8514048199512407, "grad_norm": 0.11976897716522217, "learning_rate": 3.166350325388695e-06, "loss": 0.032, "step": 142690 }, { "epoch": 2.8516046520922425, "grad_norm": 0.12776672840118408, "learning_rate": 3.162732239624855e-06, "loss": 0.0179, "step": 142700 }, { "epoch": 2.851804484233244, "grad_norm": 0.12772028148174286, "learning_rate": 3.1591161546630556e-06, "loss": 0.0265, "step": 142710 }, { "epoch": 2.852004316374246, "grad_norm": 0.11736033111810684, "learning_rate": 3.1555020706577454e-06, "loss": 0.0197, "step": 142720 }, { "epoch": 2.852204148515247, "grad_norm": 0.1011151522397995, "learning_rate": 3.1518899877633457e-06, "loss": 0.0236, "step": 142730 }, { "epoch": 2.852403980656249, "grad_norm": 0.15090426802635193, "learning_rate": 3.148279906134144e-06, "loss": 0.0306, "step": 142740 }, { "epoch": 2.85260381279725, "grad_norm": 0.12589074671268463, "learning_rate": 3.14467182592435e-06, "loss": 0.0228, "step": 142750 }, { "epoch": 2.852803644938252, "grad_norm": 0.12182388454675674, "learning_rate": 3.1410657472881024e-06, "loss": 0.0257, "step": 142760 }, { "epoch": 2.8530034770792536, "grad_norm": 0.18044719099998474, "learning_rate": 3.137461670379438e-06, "loss": 0.0272, "step": 142770 }, { "epoch": 2.853203309220255, "grad_norm": 0.13340723514556885, "learning_rate": 3.1338595953523287e-06, "loss": 0.0186, "step": 142780 }, { "epoch": 2.8534031413612566, "grad_norm": 0.10877674072980881, "learning_rate": 3.1302595223606347e-06, "loss": 0.0216, "step": 142790 }, { "epoch": 2.853602973502258, "grad_norm": 0.12749235332012177, "learning_rate": 3.126661451558155e-06, "loss": 0.0199, "step": 142800 }, { "epoch": 2.8538028056432596, "grad_norm": 0.13749054074287415, "learning_rate": 3.123065383098589e-06, "loss": 0.0225, "step": 142810 }, { "epoch": 2.8540026377842613, "grad_norm": 0.13935217261314392, "learning_rate": 3.1194713171355528e-06, "loss": 0.0235, "step": 142820 }, { "epoch": 2.854202469925263, "grad_norm": 0.16391940414905548, "learning_rate": 3.1158792538225843e-06, "loss": 0.0183, "step": 142830 }, { "epoch": 2.8544023020662643, "grad_norm": 0.15015500783920288, "learning_rate": 3.112289193313128e-06, "loss": 0.0203, "step": 142840 }, { "epoch": 2.854602134207266, "grad_norm": 0.11847997456789017, "learning_rate": 3.1087011357605335e-06, "loss": 0.0225, "step": 142850 }, { "epoch": 2.8548019663482673, "grad_norm": 0.13056373596191406, "learning_rate": 3.1051150813180775e-06, "loss": 0.0236, "step": 142860 }, { "epoch": 2.855001798489269, "grad_norm": 0.07906824350357056, "learning_rate": 3.1015310301389655e-06, "loss": 0.0173, "step": 142870 }, { "epoch": 2.8552016306302708, "grad_norm": 0.10104558616876602, "learning_rate": 3.097948982376303e-06, "loss": 0.0221, "step": 142880 }, { "epoch": 2.855401462771272, "grad_norm": 0.10444079339504242, "learning_rate": 3.0943689381830897e-06, "loss": 0.0215, "step": 142890 }, { "epoch": 2.8556012949122738, "grad_norm": 0.0837407186627388, "learning_rate": 3.0907908977122535e-06, "loss": 0.0261, "step": 142900 }, { "epoch": 2.855801127053275, "grad_norm": 0.15130019187927246, "learning_rate": 3.0872148611166774e-06, "loss": 0.023, "step": 142910 }, { "epoch": 2.8560009591942768, "grad_norm": 0.08814745396375656, "learning_rate": 3.0836408285490894e-06, "loss": 0.0219, "step": 142920 }, { "epoch": 2.8562007913352785, "grad_norm": 0.08245377987623215, "learning_rate": 3.080068800162178e-06, "loss": 0.0225, "step": 142930 }, { "epoch": 2.8564006234762798, "grad_norm": 0.09493925422430038, "learning_rate": 3.0764987761085328e-06, "loss": 0.0265, "step": 142940 }, { "epoch": 2.8566004556172815, "grad_norm": 0.07347863912582397, "learning_rate": 3.0729307565406595e-06, "loss": 0.024, "step": 142950 }, { "epoch": 2.856800287758283, "grad_norm": 0.13839493691921234, "learning_rate": 3.069364741610975e-06, "loss": 0.0202, "step": 142960 }, { "epoch": 2.8570001198992845, "grad_norm": 0.1109083741903305, "learning_rate": 3.0658007314718128e-06, "loss": 0.0264, "step": 142970 }, { "epoch": 2.857199952040286, "grad_norm": 0.127317875623703, "learning_rate": 3.062238726275435e-06, "loss": 0.0235, "step": 142980 }, { "epoch": 2.857399784181288, "grad_norm": 0.20680807530879974, "learning_rate": 3.0586787261739693e-06, "loss": 0.0196, "step": 142990 }, { "epoch": 2.857599616322289, "grad_norm": 0.32367298007011414, "learning_rate": 3.055120731319527e-06, "loss": 0.0239, "step": 143000 }, { "epoch": 2.857799448463291, "grad_norm": 0.11273596435785294, "learning_rate": 3.0515647418640935e-06, "loss": 0.0267, "step": 143010 }, { "epoch": 2.857999280604292, "grad_norm": 0.12754879891872406, "learning_rate": 3.0480107579595573e-06, "loss": 0.0242, "step": 143020 }, { "epoch": 2.858199112745294, "grad_norm": 0.125186949968338, "learning_rate": 3.044458779757753e-06, "loss": 0.0221, "step": 143030 }, { "epoch": 2.8583989448862956, "grad_norm": 0.13154973089694977, "learning_rate": 3.0409088074104085e-06, "loss": 0.0242, "step": 143040 }, { "epoch": 2.858598777027297, "grad_norm": 0.14822112023830414, "learning_rate": 3.03736084106917e-06, "loss": 0.0229, "step": 143050 }, { "epoch": 2.8587986091682986, "grad_norm": 0.08258694410324097, "learning_rate": 3.033814880885605e-06, "loss": 0.0226, "step": 143060 }, { "epoch": 2.8589984413093, "grad_norm": 0.1508326530456543, "learning_rate": 3.030270927011186e-06, "loss": 0.0309, "step": 143070 }, { "epoch": 2.8591982734503016, "grad_norm": 0.1175260916352272, "learning_rate": 3.02672897959731e-06, "loss": 0.0234, "step": 143080 }, { "epoch": 2.8593981055913034, "grad_norm": 0.1440955251455307, "learning_rate": 3.0231890387952767e-06, "loss": 0.0205, "step": 143090 }, { "epoch": 2.859597937732305, "grad_norm": 0.1414165198802948, "learning_rate": 3.0196511047563104e-06, "loss": 0.0207, "step": 143100 }, { "epoch": 2.8597977698733064, "grad_norm": 0.12573856115341187, "learning_rate": 3.016115177631551e-06, "loss": 0.0228, "step": 143110 }, { "epoch": 2.859997602014308, "grad_norm": 0.11298372596502304, "learning_rate": 3.0125812575720334e-06, "loss": 0.0257, "step": 143120 }, { "epoch": 2.8601974341553094, "grad_norm": 0.10537704080343246, "learning_rate": 3.009049344728715e-06, "loss": 0.0222, "step": 143130 }, { "epoch": 2.860397266296311, "grad_norm": 0.11820492148399353, "learning_rate": 3.0055194392524965e-06, "loss": 0.0248, "step": 143140 }, { "epoch": 2.860597098437313, "grad_norm": 0.1200128123164177, "learning_rate": 3.0019915412941524e-06, "loss": 0.0253, "step": 143150 }, { "epoch": 2.860796930578314, "grad_norm": 0.08960980921983719, "learning_rate": 2.9984656510043895e-06, "loss": 0.0183, "step": 143160 }, { "epoch": 2.860996762719316, "grad_norm": 0.10330332815647125, "learning_rate": 2.994941768533832e-06, "loss": 0.0166, "step": 143170 }, { "epoch": 2.861196594860317, "grad_norm": 0.17284484207630157, "learning_rate": 2.9914198940330095e-06, "loss": 0.0218, "step": 143180 }, { "epoch": 2.861396427001319, "grad_norm": 0.09925706684589386, "learning_rate": 2.9879000276523738e-06, "loss": 0.019, "step": 143190 }, { "epoch": 2.8615962591423205, "grad_norm": 0.1173519492149353, "learning_rate": 2.9843821695422825e-06, "loss": 0.0298, "step": 143200 }, { "epoch": 2.8617960912833222, "grad_norm": 0.18041273951530457, "learning_rate": 2.9808663198530206e-06, "loss": 0.0279, "step": 143210 }, { "epoch": 2.8619959234243235, "grad_norm": 0.15456317365169525, "learning_rate": 2.9773524787347573e-06, "loss": 0.0203, "step": 143220 }, { "epoch": 2.8621957555653252, "grad_norm": 0.11434789001941681, "learning_rate": 2.9738406463376223e-06, "loss": 0.0203, "step": 143230 }, { "epoch": 2.8623955877063265, "grad_norm": 0.1234695166349411, "learning_rate": 2.9703308228116346e-06, "loss": 0.0224, "step": 143240 }, { "epoch": 2.8625954198473282, "grad_norm": 0.09211277961730957, "learning_rate": 2.9668230083067074e-06, "loss": 0.0275, "step": 143250 }, { "epoch": 2.86279525198833, "grad_norm": 0.10743186622858047, "learning_rate": 2.963317202972693e-06, "loss": 0.0202, "step": 143260 }, { "epoch": 2.8629950841293312, "grad_norm": 0.13142910599708557, "learning_rate": 2.9598134069593607e-06, "loss": 0.0221, "step": 143270 }, { "epoch": 2.863194916270333, "grad_norm": 0.14557673037052155, "learning_rate": 2.9563116204163965e-06, "loss": 0.0304, "step": 143280 }, { "epoch": 2.8633947484113342, "grad_norm": 0.10262862592935562, "learning_rate": 2.9528118434933694e-06, "loss": 0.0246, "step": 143290 }, { "epoch": 2.863594580552336, "grad_norm": 0.10557647049427032, "learning_rate": 2.9493140763397874e-06, "loss": 0.0217, "step": 143300 }, { "epoch": 2.8637944126933377, "grad_norm": 0.1265248954296112, "learning_rate": 2.945818319105087e-06, "loss": 0.0217, "step": 143310 }, { "epoch": 2.8639942448343394, "grad_norm": 0.13562358915805817, "learning_rate": 2.942324571938576e-06, "loss": 0.0214, "step": 143320 }, { "epoch": 2.8641940769753407, "grad_norm": 0.08177264779806137, "learning_rate": 2.9388328349895132e-06, "loss": 0.0184, "step": 143330 }, { "epoch": 2.8643939091163424, "grad_norm": 0.13640914857387543, "learning_rate": 2.935343108407057e-06, "loss": 0.0255, "step": 143340 }, { "epoch": 2.8645937412573437, "grad_norm": 0.1109146922826767, "learning_rate": 2.9318553923402826e-06, "loss": 0.017, "step": 143350 }, { "epoch": 2.8647935733983454, "grad_norm": 0.11599069088697433, "learning_rate": 2.9283696869381815e-06, "loss": 0.0269, "step": 143360 }, { "epoch": 2.864993405539347, "grad_norm": 0.13679558038711548, "learning_rate": 2.924885992349652e-06, "loss": 0.0256, "step": 143370 }, { "epoch": 2.8651932376803484, "grad_norm": 0.1413857489824295, "learning_rate": 2.9214043087235187e-06, "loss": 0.0234, "step": 143380 }, { "epoch": 2.86539306982135, "grad_norm": 0.1408846378326416, "learning_rate": 2.917924636208497e-06, "loss": 0.0187, "step": 143390 }, { "epoch": 2.8655929019623514, "grad_norm": 0.11973121017217636, "learning_rate": 2.9144469749532445e-06, "loss": 0.0244, "step": 143400 }, { "epoch": 2.865792734103353, "grad_norm": 0.09804379940032959, "learning_rate": 2.9109713251063266e-06, "loss": 0.0204, "step": 143410 }, { "epoch": 2.865992566244355, "grad_norm": 0.11867068707942963, "learning_rate": 2.9074976868162073e-06, "loss": 0.0206, "step": 143420 }, { "epoch": 2.8661923983853566, "grad_norm": 0.12886351346969604, "learning_rate": 2.9040260602312685e-06, "loss": 0.0309, "step": 143430 }, { "epoch": 2.866392230526358, "grad_norm": 0.1217806413769722, "learning_rate": 2.900556445499819e-06, "loss": 0.0188, "step": 143440 }, { "epoch": 2.8665920626673596, "grad_norm": 0.1324092447757721, "learning_rate": 2.897088842770074e-06, "loss": 0.0289, "step": 143450 }, { "epoch": 2.866791894808361, "grad_norm": 0.15275925397872925, "learning_rate": 2.893623252190164e-06, "loss": 0.0258, "step": 143460 }, { "epoch": 2.8669917269493626, "grad_norm": 0.11728184670209885, "learning_rate": 2.890159673908133e-06, "loss": 0.026, "step": 143470 }, { "epoch": 2.8671915590903643, "grad_norm": 0.17556321620941162, "learning_rate": 2.8866981080719336e-06, "loss": 0.0264, "step": 143480 }, { "epoch": 2.8673913912313655, "grad_norm": 0.13588891923427582, "learning_rate": 2.8832385548294427e-06, "loss": 0.0217, "step": 143490 }, { "epoch": 2.8675912233723673, "grad_norm": 0.11822541058063507, "learning_rate": 2.8797810143284475e-06, "loss": 0.0166, "step": 143500 }, { "epoch": 2.8677910555133685, "grad_norm": 0.14937089383602142, "learning_rate": 2.876325486716647e-06, "loss": 0.0217, "step": 143510 }, { "epoch": 2.8679908876543703, "grad_norm": 0.10323217511177063, "learning_rate": 2.8728719721416498e-06, "loss": 0.0206, "step": 143520 }, { "epoch": 2.868190719795372, "grad_norm": 0.09750957041978836, "learning_rate": 2.8694204707509776e-06, "loss": 0.0209, "step": 143530 }, { "epoch": 2.8683905519363737, "grad_norm": 0.09926308691501617, "learning_rate": 2.8659709826920955e-06, "loss": 0.0259, "step": 143540 }, { "epoch": 2.868590384077375, "grad_norm": 0.09699169546365738, "learning_rate": 2.862523508112336e-06, "loss": 0.0242, "step": 143550 }, { "epoch": 2.8687902162183767, "grad_norm": 0.15848681330680847, "learning_rate": 2.859078047158981e-06, "loss": 0.0246, "step": 143560 }, { "epoch": 2.868990048359378, "grad_norm": 0.1491224616765976, "learning_rate": 2.855634599979212e-06, "loss": 0.0263, "step": 143570 }, { "epoch": 2.8691898805003797, "grad_norm": 0.12072066217660904, "learning_rate": 2.852193166720124e-06, "loss": 0.0237, "step": 143580 }, { "epoch": 2.8693897126413814, "grad_norm": 0.11797915399074554, "learning_rate": 2.848753747528726e-06, "loss": 0.0287, "step": 143590 }, { "epoch": 2.8695895447823827, "grad_norm": 0.11538955569267273, "learning_rate": 2.845316342551957e-06, "loss": 0.0216, "step": 143600 }, { "epoch": 2.8697893769233844, "grad_norm": 0.1327606439590454, "learning_rate": 2.8418809519366485e-06, "loss": 0.0225, "step": 143610 }, { "epoch": 2.8699892090643857, "grad_norm": 0.11681971698999405, "learning_rate": 2.8384475758295393e-06, "loss": 0.0225, "step": 143620 }, { "epoch": 2.8701890412053874, "grad_norm": 0.13138554990291595, "learning_rate": 2.835016214377323e-06, "loss": 0.0249, "step": 143630 }, { "epoch": 2.870388873346389, "grad_norm": 0.11818437278270721, "learning_rate": 2.831586867726577e-06, "loss": 0.029, "step": 143640 }, { "epoch": 2.870588705487391, "grad_norm": 0.10801970213651657, "learning_rate": 2.828159536023778e-06, "loss": 0.024, "step": 143650 }, { "epoch": 2.870788537628392, "grad_norm": 0.254072904586792, "learning_rate": 2.824734219415337e-06, "loss": 0.0282, "step": 143660 }, { "epoch": 2.870988369769394, "grad_norm": 0.12220720201730728, "learning_rate": 2.821310918047598e-06, "loss": 0.0248, "step": 143670 }, { "epoch": 2.871188201910395, "grad_norm": 0.16361556947231293, "learning_rate": 2.8178896320667892e-06, "loss": 0.0199, "step": 143680 }, { "epoch": 2.871388034051397, "grad_norm": 0.09021604061126709, "learning_rate": 2.814470361619054e-06, "loss": 0.0226, "step": 143690 }, { "epoch": 2.8715878661923986, "grad_norm": 0.13829989731311798, "learning_rate": 2.8110531068504643e-06, "loss": 0.0191, "step": 143700 }, { "epoch": 2.8717876983334, "grad_norm": 0.16641606390476227, "learning_rate": 2.807637867906998e-06, "loss": 0.0179, "step": 143710 }, { "epoch": 2.8719875304744016, "grad_norm": 0.14382435381412506, "learning_rate": 2.8042246449345444e-06, "loss": 0.028, "step": 143720 }, { "epoch": 2.872187362615403, "grad_norm": 0.09499069303274155, "learning_rate": 2.8008134380789085e-06, "loss": 0.021, "step": 143730 }, { "epoch": 2.8723871947564046, "grad_norm": 0.10330840945243835, "learning_rate": 2.7974042474858187e-06, "loss": 0.0194, "step": 143740 }, { "epoch": 2.8725870268974063, "grad_norm": 0.12104672193527222, "learning_rate": 2.793997073300908e-06, "loss": 0.0245, "step": 143750 }, { "epoch": 2.8727868590384076, "grad_norm": 0.17184455692768097, "learning_rate": 2.7905919156697214e-06, "loss": 0.0286, "step": 143760 }, { "epoch": 2.8729866911794093, "grad_norm": 0.10741442441940308, "learning_rate": 2.7871887747377203e-06, "loss": 0.0266, "step": 143770 }, { "epoch": 2.8731865233204106, "grad_norm": 0.13028384745121002, "learning_rate": 2.783787650650288e-06, "loss": 0.023, "step": 143780 }, { "epoch": 2.8733863554614123, "grad_norm": 0.11244960129261017, "learning_rate": 2.780388543552703e-06, "loss": 0.024, "step": 143790 }, { "epoch": 2.873586187602414, "grad_norm": 0.11456074565649033, "learning_rate": 2.7769914535901665e-06, "loss": 0.0242, "step": 143800 }, { "epoch": 2.8737860197434157, "grad_norm": 0.13744181394577026, "learning_rate": 2.773596380907817e-06, "loss": 0.0216, "step": 143810 }, { "epoch": 2.873985851884417, "grad_norm": 0.15120171010494232, "learning_rate": 2.7702033256506664e-06, "loss": 0.02, "step": 143820 }, { "epoch": 2.8741856840254187, "grad_norm": 0.1455260068178177, "learning_rate": 2.766812287963666e-06, "loss": 0.016, "step": 143830 }, { "epoch": 2.87438551616642, "grad_norm": 0.13966406881809235, "learning_rate": 2.7634232679916718e-06, "loss": 0.0257, "step": 143840 }, { "epoch": 2.8745853483074217, "grad_norm": 0.07538682222366333, "learning_rate": 2.760036265879462e-06, "loss": 0.0219, "step": 143850 }, { "epoch": 2.8747851804484235, "grad_norm": 0.12775304913520813, "learning_rate": 2.7566512817717215e-06, "loss": 0.0238, "step": 143860 }, { "epoch": 2.8749850125894247, "grad_norm": 0.11025752127170563, "learning_rate": 2.7532683158130457e-06, "loss": 0.0228, "step": 143870 }, { "epoch": 2.8751848447304265, "grad_norm": 0.11737509816884995, "learning_rate": 2.749887368147963e-06, "loss": 0.0273, "step": 143880 }, { "epoch": 2.8753846768714277, "grad_norm": 0.08244863897562027, "learning_rate": 2.746508438920875e-06, "loss": 0.028, "step": 143890 }, { "epoch": 2.8755845090124295, "grad_norm": 0.09456358104944229, "learning_rate": 2.743131528276144e-06, "loss": 0.0316, "step": 143900 }, { "epoch": 2.875784341153431, "grad_norm": 0.13176874816417694, "learning_rate": 2.7397566363580318e-06, "loss": 0.0271, "step": 143910 }, { "epoch": 2.875984173294433, "grad_norm": 0.10448875278234482, "learning_rate": 2.7363837633106847e-06, "loss": 0.0254, "step": 143920 }, { "epoch": 2.876184005435434, "grad_norm": 0.13423268496990204, "learning_rate": 2.733012909278193e-06, "loss": 0.0279, "step": 143930 }, { "epoch": 2.876383837576436, "grad_norm": 0.14725050330162048, "learning_rate": 2.729644074404575e-06, "loss": 0.0258, "step": 143940 }, { "epoch": 2.876583669717437, "grad_norm": 0.08362586796283722, "learning_rate": 2.7262772588337095e-06, "loss": 0.0191, "step": 143950 }, { "epoch": 2.876783501858439, "grad_norm": 0.08608462661504745, "learning_rate": 2.722912462709437e-06, "loss": 0.0209, "step": 143960 }, { "epoch": 2.8769833339994406, "grad_norm": 0.17720100283622742, "learning_rate": 2.719549686175499e-06, "loss": 0.0267, "step": 143970 }, { "epoch": 2.877183166140442, "grad_norm": 0.08463972061872482, "learning_rate": 2.716188929375535e-06, "loss": 0.0284, "step": 143980 }, { "epoch": 2.8773829982814436, "grad_norm": 0.13130322098731995, "learning_rate": 2.7128301924531198e-06, "loss": 0.0297, "step": 143990 }, { "epoch": 2.877582830422445, "grad_norm": 0.13094380497932434, "learning_rate": 2.709473475551727e-06, "loss": 0.0208, "step": 144000 }, { "epoch": 2.8777826625634466, "grad_norm": 0.10204519331455231, "learning_rate": 2.7061187788147592e-06, "loss": 0.0197, "step": 144010 }, { "epoch": 2.8779824947044483, "grad_norm": 0.11725448817014694, "learning_rate": 2.7027661023855066e-06, "loss": 0.0212, "step": 144020 }, { "epoch": 2.87818232684545, "grad_norm": 0.09217662364244461, "learning_rate": 2.6994154464071995e-06, "loss": 0.0181, "step": 144030 }, { "epoch": 2.8783821589864513, "grad_norm": 0.10061226785182953, "learning_rate": 2.696066811022979e-06, "loss": 0.0238, "step": 144040 }, { "epoch": 2.878581991127453, "grad_norm": 0.19604185223579407, "learning_rate": 2.6927201963758807e-06, "loss": 0.0254, "step": 144050 }, { "epoch": 2.8787818232684543, "grad_norm": 0.1171504557132721, "learning_rate": 2.6893756026088624e-06, "loss": 0.028, "step": 144060 }, { "epoch": 2.878981655409456, "grad_norm": 0.08490884304046631, "learning_rate": 2.686033029864804e-06, "loss": 0.02, "step": 144070 }, { "epoch": 2.8791814875504578, "grad_norm": 0.14926505088806152, "learning_rate": 2.6826924782865082e-06, "loss": 0.0255, "step": 144080 }, { "epoch": 2.879381319691459, "grad_norm": 0.0817667692899704, "learning_rate": 2.6793539480166607e-06, "loss": 0.0204, "step": 144090 }, { "epoch": 2.8795811518324608, "grad_norm": 0.11143916100263596, "learning_rate": 2.676017439197881e-06, "loss": 0.0244, "step": 144100 }, { "epoch": 2.879780983973462, "grad_norm": 0.1670742779970169, "learning_rate": 2.672682951972699e-06, "loss": 0.0255, "step": 144110 }, { "epoch": 2.8799808161144638, "grad_norm": 0.11906179040670395, "learning_rate": 2.669350486483557e-06, "loss": 0.022, "step": 144120 }, { "epoch": 2.8801806482554655, "grad_norm": 0.12729232013225555, "learning_rate": 2.6660200428728133e-06, "loss": 0.0229, "step": 144130 }, { "epoch": 2.880380480396467, "grad_norm": 0.1310487687587738, "learning_rate": 2.662691621282737e-06, "loss": 0.0277, "step": 144140 }, { "epoch": 2.8805803125374685, "grad_norm": 0.1367788016796112, "learning_rate": 2.6593652218555196e-06, "loss": 0.021, "step": 144150 }, { "epoch": 2.88078014467847, "grad_norm": 0.12021464109420776, "learning_rate": 2.656040844733243e-06, "loss": 0.0216, "step": 144160 }, { "epoch": 2.8809799768194715, "grad_norm": 0.11202831566333771, "learning_rate": 2.6527184900579315e-06, "loss": 0.0244, "step": 144170 }, { "epoch": 2.881179808960473, "grad_norm": 0.1501764953136444, "learning_rate": 2.6493981579715167e-06, "loss": 0.024, "step": 144180 }, { "epoch": 2.881379641101475, "grad_norm": 0.10128781944513321, "learning_rate": 2.646079848615812e-06, "loss": 0.0216, "step": 144190 }, { "epoch": 2.881579473242476, "grad_norm": 0.13770990073680878, "learning_rate": 2.642763562132583e-06, "loss": 0.0231, "step": 144200 }, { "epoch": 2.881779305383478, "grad_norm": 0.12238963693380356, "learning_rate": 2.63944929866351e-06, "loss": 0.0287, "step": 144210 }, { "epoch": 2.881979137524479, "grad_norm": 0.13100585341453552, "learning_rate": 2.636137058350152e-06, "loss": 0.0239, "step": 144220 }, { "epoch": 2.882178969665481, "grad_norm": 0.09789098054170609, "learning_rate": 2.6328268413340074e-06, "loss": 0.0223, "step": 144230 }, { "epoch": 2.8823788018064826, "grad_norm": 0.10870250314474106, "learning_rate": 2.6295186477564847e-06, "loss": 0.0168, "step": 144240 }, { "epoch": 2.8825786339474844, "grad_norm": 0.15345607697963715, "learning_rate": 2.626212477758905e-06, "loss": 0.0242, "step": 144250 }, { "epoch": 2.8827784660884856, "grad_norm": 0.08794289827346802, "learning_rate": 2.6229083314824985e-06, "loss": 0.0197, "step": 144260 }, { "epoch": 2.8829782982294874, "grad_norm": 0.1569211781024933, "learning_rate": 2.6196062090684147e-06, "loss": 0.0251, "step": 144270 }, { "epoch": 2.8831781303704886, "grad_norm": 0.1203957349061966, "learning_rate": 2.616306110657718e-06, "loss": 0.0203, "step": 144280 }, { "epoch": 2.8833779625114904, "grad_norm": 0.11109217256307602, "learning_rate": 2.6130080363913624e-06, "loss": 0.0256, "step": 144290 }, { "epoch": 2.883577794652492, "grad_norm": 0.1474146544933319, "learning_rate": 2.6097119864102628e-06, "loss": 0.0219, "step": 144300 }, { "epoch": 2.8837776267934934, "grad_norm": 0.12263903021812439, "learning_rate": 2.606417960855212e-06, "loss": 0.0225, "step": 144310 }, { "epoch": 2.883977458934495, "grad_norm": 0.1405235379934311, "learning_rate": 2.6031259598669144e-06, "loss": 0.0219, "step": 144320 }, { "epoch": 2.8841772910754964, "grad_norm": 0.10970321297645569, "learning_rate": 2.599835983586002e-06, "loss": 0.0238, "step": 144330 }, { "epoch": 2.884377123216498, "grad_norm": 0.0815156027674675, "learning_rate": 2.5965480321530235e-06, "loss": 0.0196, "step": 144340 }, { "epoch": 2.8845769553575, "grad_norm": 0.18293409049510956, "learning_rate": 2.5932621057084273e-06, "loss": 0.0255, "step": 144350 }, { "epoch": 2.8847767874985015, "grad_norm": 0.09018415957689285, "learning_rate": 2.5899782043925904e-06, "loss": 0.0172, "step": 144360 }, { "epoch": 2.884976619639503, "grad_norm": 0.13512209057807922, "learning_rate": 2.5866963283457834e-06, "loss": 0.0286, "step": 144370 }, { "epoch": 2.8851764517805045, "grad_norm": 0.1321825087070465, "learning_rate": 2.5834164777082115e-06, "loss": 0.0237, "step": 144380 }, { "epoch": 2.885376283921506, "grad_norm": 0.08668485283851624, "learning_rate": 2.580138652619979e-06, "loss": 0.0227, "step": 144390 }, { "epoch": 2.8855761160625075, "grad_norm": 0.10993081331253052, "learning_rate": 2.5768628532211125e-06, "loss": 0.0276, "step": 144400 }, { "epoch": 2.8857759482035092, "grad_norm": 0.11817373335361481, "learning_rate": 2.5735890796515506e-06, "loss": 0.0216, "step": 144410 }, { "epoch": 2.8859757803445105, "grad_norm": 0.08204551041126251, "learning_rate": 2.570317332051131e-06, "loss": 0.0222, "step": 144420 }, { "epoch": 2.8861756124855122, "grad_norm": 0.12861856818199158, "learning_rate": 2.5670476105596197e-06, "loss": 0.0291, "step": 144430 }, { "epoch": 2.8863754446265135, "grad_norm": 0.09130217134952545, "learning_rate": 2.563779915316705e-06, "loss": 0.0233, "step": 144440 }, { "epoch": 2.8865752767675152, "grad_norm": 0.10742152482271194, "learning_rate": 2.56051424646197e-06, "loss": 0.0251, "step": 144450 }, { "epoch": 2.886775108908517, "grad_norm": 0.1279938817024231, "learning_rate": 2.557250604134914e-06, "loss": 0.0256, "step": 144460 }, { "epoch": 2.8869749410495182, "grad_norm": 0.105233334004879, "learning_rate": 2.5539889884749478e-06, "loss": 0.0174, "step": 144470 }, { "epoch": 2.88717477319052, "grad_norm": 0.11906212568283081, "learning_rate": 2.5507293996214266e-06, "loss": 0.023, "step": 144480 }, { "epoch": 2.8873746053315217, "grad_norm": 0.1003839448094368, "learning_rate": 2.547471837713572e-06, "loss": 0.0242, "step": 144490 }, { "epoch": 2.887574437472523, "grad_norm": 0.10162264853715897, "learning_rate": 2.5442163028905507e-06, "loss": 0.0257, "step": 144500 }, { "epoch": 2.8877742696135247, "grad_norm": 0.09024051576852798, "learning_rate": 2.540962795291424e-06, "loss": 0.018, "step": 144510 }, { "epoch": 2.8879741017545264, "grad_norm": 0.07553713023662567, "learning_rate": 2.537711315055186e-06, "loss": 0.0242, "step": 144520 }, { "epoch": 2.8881739338955277, "grad_norm": 0.1020333394408226, "learning_rate": 2.534461862320725e-06, "loss": 0.0158, "step": 144530 }, { "epoch": 2.8883737660365294, "grad_norm": 0.09375613182783127, "learning_rate": 2.531214437226859e-06, "loss": 0.023, "step": 144540 }, { "epoch": 2.8885735981775307, "grad_norm": 0.11342581361532211, "learning_rate": 2.5279690399123144e-06, "loss": 0.0203, "step": 144550 }, { "epoch": 2.8887734303185324, "grad_norm": 0.12639781832695007, "learning_rate": 2.524725670515715e-06, "loss": 0.0193, "step": 144560 }, { "epoch": 2.888973262459534, "grad_norm": 0.11303005367517471, "learning_rate": 2.521484329175622e-06, "loss": 0.0196, "step": 144570 }, { "epoch": 2.8891730946005354, "grad_norm": 0.09804577380418777, "learning_rate": 2.5182450160305073e-06, "loss": 0.0301, "step": 144580 }, { "epoch": 2.889372926741537, "grad_norm": 0.11052308976650238, "learning_rate": 2.515007731218727e-06, "loss": 0.0226, "step": 144590 }, { "epoch": 2.8895727588825384, "grad_norm": 0.09704697132110596, "learning_rate": 2.511772474878582e-06, "loss": 0.0194, "step": 144600 }, { "epoch": 2.88977259102354, "grad_norm": 0.1395363062620163, "learning_rate": 2.508539247148289e-06, "loss": 0.0283, "step": 144610 }, { "epoch": 2.889972423164542, "grad_norm": 0.12377314269542694, "learning_rate": 2.5053080481659498e-06, "loss": 0.0172, "step": 144620 }, { "epoch": 2.8901722553055436, "grad_norm": 0.09723369032144547, "learning_rate": 2.5020788780696024e-06, "loss": 0.0212, "step": 144630 }, { "epoch": 2.890372087446545, "grad_norm": 0.09364766627550125, "learning_rate": 2.498851736997182e-06, "loss": 0.0237, "step": 144640 }, { "epoch": 2.8905719195875466, "grad_norm": 0.07710868865251541, "learning_rate": 2.495626625086561e-06, "loss": 0.018, "step": 144650 }, { "epoch": 2.890771751728548, "grad_norm": 0.13667574524879456, "learning_rate": 2.4924035424754968e-06, "loss": 0.0247, "step": 144660 }, { "epoch": 2.8909715838695496, "grad_norm": 0.09891945123672485, "learning_rate": 2.489182489301678e-06, "loss": 0.0288, "step": 144670 }, { "epoch": 2.8911714160105513, "grad_norm": 0.11804093420505524, "learning_rate": 2.4859634657027175e-06, "loss": 0.0307, "step": 144680 }, { "epoch": 2.8913712481515526, "grad_norm": 0.13198277354240417, "learning_rate": 2.482746471816089e-06, "loss": 0.0218, "step": 144690 }, { "epoch": 2.8915710802925543, "grad_norm": 0.08930914849042892, "learning_rate": 2.4795315077792536e-06, "loss": 0.0267, "step": 144700 }, { "epoch": 2.8917709124335556, "grad_norm": 0.1560000628232956, "learning_rate": 2.4763185737295358e-06, "loss": 0.023, "step": 144710 }, { "epoch": 2.8919707445745573, "grad_norm": 0.16759967803955078, "learning_rate": 2.473107669804181e-06, "loss": 0.0242, "step": 144720 }, { "epoch": 2.892170576715559, "grad_norm": 0.08031988888978958, "learning_rate": 2.469898796140352e-06, "loss": 0.0194, "step": 144730 }, { "epoch": 2.8923704088565607, "grad_norm": 0.11419704556465149, "learning_rate": 2.466691952875139e-06, "loss": 0.022, "step": 144740 }, { "epoch": 2.892570240997562, "grad_norm": 0.1444379687309265, "learning_rate": 2.463487140145515e-06, "loss": 0.0223, "step": 144750 }, { "epoch": 2.8927700731385637, "grad_norm": 0.12256241589784622, "learning_rate": 2.4602843580883995e-06, "loss": 0.026, "step": 144760 }, { "epoch": 2.892969905279565, "grad_norm": 0.14701756834983826, "learning_rate": 2.457083606840599e-06, "loss": 0.0244, "step": 144770 }, { "epoch": 2.8931697374205667, "grad_norm": 0.10423426330089569, "learning_rate": 2.4538848865388543e-06, "loss": 0.0219, "step": 144780 }, { "epoch": 2.8933695695615684, "grad_norm": 0.10442736744880676, "learning_rate": 2.4506881973197947e-06, "loss": 0.0216, "step": 144790 }, { "epoch": 2.8935694017025697, "grad_norm": 0.10917571932077408, "learning_rate": 2.4474935393199894e-06, "loss": 0.0269, "step": 144800 }, { "epoch": 2.8937692338435714, "grad_norm": 0.10498934239149094, "learning_rate": 2.4443009126759064e-06, "loss": 0.0295, "step": 144810 }, { "epoch": 2.8939690659845727, "grad_norm": 0.11816957592964172, "learning_rate": 2.4411103175239202e-06, "loss": 0.0309, "step": 144820 }, { "epoch": 2.8941688981255744, "grad_norm": 0.4186573326587677, "learning_rate": 2.4379217540003273e-06, "loss": 0.0186, "step": 144830 }, { "epoch": 2.894368730266576, "grad_norm": 0.111956886947155, "learning_rate": 2.434735222241352e-06, "loss": 0.0264, "step": 144840 }, { "epoch": 2.894568562407578, "grad_norm": 0.12759540975093842, "learning_rate": 2.431550722383108e-06, "loss": 0.0261, "step": 144850 }, { "epoch": 2.894768394548579, "grad_norm": 0.10810090601444244, "learning_rate": 2.4283682545616304e-06, "loss": 0.0216, "step": 144860 }, { "epoch": 2.894968226689581, "grad_norm": 0.15071377158164978, "learning_rate": 2.4251878189128607e-06, "loss": 0.027, "step": 144870 }, { "epoch": 2.895168058830582, "grad_norm": 0.10773256421089172, "learning_rate": 2.422009415572679e-06, "loss": 0.0262, "step": 144880 }, { "epoch": 2.895367890971584, "grad_norm": 0.152613565325737, "learning_rate": 2.4188330446768493e-06, "loss": 0.0245, "step": 144890 }, { "epoch": 2.8955677231125856, "grad_norm": 0.13126204907894135, "learning_rate": 2.4156587063610625e-06, "loss": 0.0302, "step": 144900 }, { "epoch": 2.895767555253587, "grad_norm": 0.10174966603517532, "learning_rate": 2.412486400760916e-06, "loss": 0.0242, "step": 144910 }, { "epoch": 2.8959673873945886, "grad_norm": 0.13120874762535095, "learning_rate": 2.4093161280119347e-06, "loss": 0.0259, "step": 144920 }, { "epoch": 2.89616721953559, "grad_norm": 0.12167290598154068, "learning_rate": 2.4061478882495382e-06, "loss": 0.0308, "step": 144930 }, { "epoch": 2.8963670516765916, "grad_norm": 0.10948402434587479, "learning_rate": 2.4029816816090735e-06, "loss": 0.0193, "step": 144940 }, { "epoch": 2.8965668838175933, "grad_norm": 0.13622167706489563, "learning_rate": 2.399817508225799e-06, "loss": 0.0243, "step": 144950 }, { "epoch": 2.896766715958595, "grad_norm": 0.12113989889621735, "learning_rate": 2.3966553682348625e-06, "loss": 0.0303, "step": 144960 }, { "epoch": 2.8969665480995963, "grad_norm": 0.12196283042430878, "learning_rate": 2.3934952617713612e-06, "loss": 0.0248, "step": 144970 }, { "epoch": 2.897166380240598, "grad_norm": 0.11063626408576965, "learning_rate": 2.3903371889702987e-06, "loss": 0.0221, "step": 144980 }, { "epoch": 2.8973662123815993, "grad_norm": 0.1440901905298233, "learning_rate": 2.3871811499665606e-06, "loss": 0.0224, "step": 144990 }, { "epoch": 2.897566044522601, "grad_norm": 0.11786268651485443, "learning_rate": 2.384027144894979e-06, "loss": 0.0293, "step": 145000 }, { "epoch": 2.8977658766636027, "grad_norm": 0.10898523777723312, "learning_rate": 2.380875173890279e-06, "loss": 0.0182, "step": 145010 }, { "epoch": 2.897965708804604, "grad_norm": 0.17735502123832703, "learning_rate": 2.3777252370871195e-06, "loss": 0.0261, "step": 145020 }, { "epoch": 2.8981655409456057, "grad_norm": 0.1641787439584732, "learning_rate": 2.3745773346200483e-06, "loss": 0.025, "step": 145030 }, { "epoch": 2.898365373086607, "grad_norm": 0.1028374582529068, "learning_rate": 2.3714314666235415e-06, "loss": 0.0226, "step": 145040 }, { "epoch": 2.8985652052276087, "grad_norm": 0.1409441977739334, "learning_rate": 2.3682876332319915e-06, "loss": 0.0208, "step": 145050 }, { "epoch": 2.8987650373686105, "grad_norm": 0.15462177991867065, "learning_rate": 2.3651458345796853e-06, "loss": 0.0275, "step": 145060 }, { "epoch": 2.898964869509612, "grad_norm": 0.11365392804145813, "learning_rate": 2.3620060708008428e-06, "loss": 0.0294, "step": 145070 }, { "epoch": 2.8991647016506135, "grad_norm": 0.08575664460659027, "learning_rate": 2.3588683420295963e-06, "loss": 0.0252, "step": 145080 }, { "epoch": 2.899364533791615, "grad_norm": 0.1263335794210434, "learning_rate": 2.3557326483999664e-06, "loss": 0.0248, "step": 145090 }, { "epoch": 2.8995643659326165, "grad_norm": 0.11287572979927063, "learning_rate": 2.3525989900459067e-06, "loss": 0.0318, "step": 145100 }, { "epoch": 2.899764198073618, "grad_norm": 0.16765201091766357, "learning_rate": 2.349467367101299e-06, "loss": 0.0217, "step": 145110 }, { "epoch": 2.89996403021462, "grad_norm": 0.11474329978227615, "learning_rate": 2.346337779699903e-06, "loss": 0.0262, "step": 145120 }, { "epoch": 2.900163862355621, "grad_norm": 0.13017641007900238, "learning_rate": 2.343210227975412e-06, "loss": 0.0349, "step": 145130 }, { "epoch": 2.900363694496623, "grad_norm": 0.12428656965494156, "learning_rate": 2.3400847120614356e-06, "loss": 0.0255, "step": 145140 }, { "epoch": 2.900563526637624, "grad_norm": 0.10892200469970703, "learning_rate": 2.3369612320914836e-06, "loss": 0.0241, "step": 145150 }, { "epoch": 2.900763358778626, "grad_norm": 0.10854271054267883, "learning_rate": 2.3338397881989883e-06, "loss": 0.0265, "step": 145160 }, { "epoch": 2.9009631909196276, "grad_norm": 0.088678739964962, "learning_rate": 2.330720380517293e-06, "loss": 0.0199, "step": 145170 }, { "epoch": 2.9011630230606293, "grad_norm": 0.1157708615064621, "learning_rate": 2.3276030091796576e-06, "loss": 0.0202, "step": 145180 }, { "epoch": 2.9013628552016306, "grad_norm": 0.09686696529388428, "learning_rate": 2.3244876743192314e-06, "loss": 0.0256, "step": 145190 }, { "epoch": 2.9015626873426323, "grad_norm": 0.16377536952495575, "learning_rate": 2.321374376069119e-06, "loss": 0.0301, "step": 145200 }, { "epoch": 2.9017625194836336, "grad_norm": 0.1340765506029129, "learning_rate": 2.318263114562308e-06, "loss": 0.0272, "step": 145210 }, { "epoch": 2.9019623516246353, "grad_norm": 0.1448458582162857, "learning_rate": 2.315153889931693e-06, "loss": 0.0256, "step": 145220 }, { "epoch": 2.902162183765637, "grad_norm": 0.10467325896024704, "learning_rate": 2.3120467023101e-06, "loss": 0.0226, "step": 145230 }, { "epoch": 2.9023620159066383, "grad_norm": 0.17029082775115967, "learning_rate": 2.3089415518302736e-06, "loss": 0.0234, "step": 145240 }, { "epoch": 2.90256184804764, "grad_norm": 0.08292963355779648, "learning_rate": 2.305838438624858e-06, "loss": 0.0247, "step": 145250 }, { "epoch": 2.9027616801886413, "grad_norm": 0.13804171979427338, "learning_rate": 2.3027373628264016e-06, "loss": 0.0192, "step": 145260 }, { "epoch": 2.902961512329643, "grad_norm": 0.11951971054077148, "learning_rate": 2.2996383245673826e-06, "loss": 0.0158, "step": 145270 }, { "epoch": 2.903161344470645, "grad_norm": 0.0940200611948967, "learning_rate": 2.296541323980189e-06, "loss": 0.0287, "step": 145280 }, { "epoch": 2.903361176611646, "grad_norm": 0.10116726160049438, "learning_rate": 2.2934463611971104e-06, "loss": 0.0272, "step": 145290 }, { "epoch": 2.9035610087526478, "grad_norm": 0.13813088834285736, "learning_rate": 2.2903534363503678e-06, "loss": 0.0239, "step": 145300 }, { "epoch": 2.9037608408936495, "grad_norm": 0.09670281410217285, "learning_rate": 2.2872625495720777e-06, "loss": 0.0267, "step": 145310 }, { "epoch": 2.9039606730346508, "grad_norm": 0.08557304739952087, "learning_rate": 2.2841737009942854e-06, "loss": 0.0209, "step": 145320 }, { "epoch": 2.9041605051756525, "grad_norm": 0.1073211133480072, "learning_rate": 2.2810868907489293e-06, "loss": 0.023, "step": 145330 }, { "epoch": 2.904360337316654, "grad_norm": 0.08132769912481308, "learning_rate": 2.278002118967881e-06, "loss": 0.0231, "step": 145340 }, { "epoch": 2.9045601694576555, "grad_norm": 0.09808437526226044, "learning_rate": 2.2749193857829256e-06, "loss": 0.0225, "step": 145350 }, { "epoch": 2.904760001598657, "grad_norm": 0.1460956186056137, "learning_rate": 2.2718386913257285e-06, "loss": 0.0207, "step": 145360 }, { "epoch": 2.9049598337396585, "grad_norm": 0.1203080341219902, "learning_rate": 2.268760035727896e-06, "loss": 0.0336, "step": 145370 }, { "epoch": 2.90515966588066, "grad_norm": 0.15239796042442322, "learning_rate": 2.2656834191209676e-06, "loss": 0.0266, "step": 145380 }, { "epoch": 2.905359498021662, "grad_norm": 0.11436508595943451, "learning_rate": 2.262608841636343e-06, "loss": 0.0289, "step": 145390 }, { "epoch": 2.905559330162663, "grad_norm": 0.09973254799842834, "learning_rate": 2.2595363034053686e-06, "loss": 0.0171, "step": 145400 }, { "epoch": 2.905759162303665, "grad_norm": 0.14244970679283142, "learning_rate": 2.25646580455931e-06, "loss": 0.0175, "step": 145410 }, { "epoch": 2.905958994444666, "grad_norm": 0.13856156170368195, "learning_rate": 2.253397345229319e-06, "loss": 0.0231, "step": 145420 }, { "epoch": 2.906158826585668, "grad_norm": 0.11974650621414185, "learning_rate": 2.2503309255464787e-06, "loss": 0.0208, "step": 145430 }, { "epoch": 2.9063586587266697, "grad_norm": 0.20380310714244843, "learning_rate": 2.247266545641785e-06, "loss": 0.0238, "step": 145440 }, { "epoch": 2.9065584908676714, "grad_norm": 0.07776081562042236, "learning_rate": 2.244204205646144e-06, "loss": 0.0235, "step": 145450 }, { "epoch": 2.9067583230086727, "grad_norm": 0.10266225785017014, "learning_rate": 2.241143905690357e-06, "loss": 0.0218, "step": 145460 }, { "epoch": 2.9069581551496744, "grad_norm": 0.11067748814821243, "learning_rate": 2.2380856459051747e-06, "loss": 0.0314, "step": 145470 }, { "epoch": 2.9071579872906756, "grad_norm": 0.11697530746459961, "learning_rate": 2.2350294264212367e-06, "loss": 0.028, "step": 145480 }, { "epoch": 2.9073578194316774, "grad_norm": 0.16057899594306946, "learning_rate": 2.2319752473690947e-06, "loss": 0.0253, "step": 145490 }, { "epoch": 2.907557651572679, "grad_norm": 0.11822797358036041, "learning_rate": 2.228923108879205e-06, "loss": 0.0182, "step": 145500 }, { "epoch": 2.9077574837136804, "grad_norm": 0.1126047819852829, "learning_rate": 2.2258730110819748e-06, "loss": 0.0326, "step": 145510 }, { "epoch": 2.907957315854682, "grad_norm": 0.13654309511184692, "learning_rate": 2.222824954107677e-06, "loss": 0.0227, "step": 145520 }, { "epoch": 2.9081571479956834, "grad_norm": 0.07906608283519745, "learning_rate": 2.2197789380865307e-06, "loss": 0.0228, "step": 145530 }, { "epoch": 2.908356980136685, "grad_norm": 0.14458969235420227, "learning_rate": 2.216734963148653e-06, "loss": 0.0258, "step": 145540 }, { "epoch": 2.908556812277687, "grad_norm": 0.10366418212652206, "learning_rate": 2.21369302942408e-06, "loss": 0.0275, "step": 145550 }, { "epoch": 2.9087566444186885, "grad_norm": 0.09192979335784912, "learning_rate": 2.2106531370427506e-06, "loss": 0.0353, "step": 145560 }, { "epoch": 2.90895647655969, "grad_norm": 0.13052251935005188, "learning_rate": 2.2076152861345288e-06, "loss": 0.0231, "step": 145570 }, { "epoch": 2.9091563087006915, "grad_norm": 0.11945362389087677, "learning_rate": 2.204579476829188e-06, "loss": 0.0328, "step": 145580 }, { "epoch": 2.909356140841693, "grad_norm": 0.11654576659202576, "learning_rate": 2.2015457092563974e-06, "loss": 0.0333, "step": 145590 }, { "epoch": 2.9095559729826945, "grad_norm": 0.09959519654512405, "learning_rate": 2.1985139835457745e-06, "loss": 0.0295, "step": 145600 }, { "epoch": 2.9097558051236962, "grad_norm": 0.09286818653345108, "learning_rate": 2.195484299826822e-06, "loss": 0.0237, "step": 145610 }, { "epoch": 2.9099556372646975, "grad_norm": 0.12575018405914307, "learning_rate": 2.1924566582289584e-06, "loss": 0.028, "step": 145620 }, { "epoch": 2.9101554694056992, "grad_norm": 0.09158404916524887, "learning_rate": 2.1894310588815194e-06, "loss": 0.0269, "step": 145630 }, { "epoch": 2.9103553015467005, "grad_norm": 0.12319949269294739, "learning_rate": 2.1864075019137507e-06, "loss": 0.0201, "step": 145640 }, { "epoch": 2.9105551336877022, "grad_norm": 0.08587297052145004, "learning_rate": 2.183385987454828e-06, "loss": 0.026, "step": 145650 }, { "epoch": 2.910754965828704, "grad_norm": 0.12083832919597626, "learning_rate": 2.180366515633808e-06, "loss": 0.0317, "step": 145660 }, { "epoch": 2.9109547979697057, "grad_norm": 0.10485625267028809, "learning_rate": 2.177349086579683e-06, "loss": 0.0249, "step": 145670 }, { "epoch": 2.911154630110707, "grad_norm": 0.06020165979862213, "learning_rate": 2.1743337004213493e-06, "loss": 0.0245, "step": 145680 }, { "epoch": 2.9113544622517087, "grad_norm": 0.12452924996614456, "learning_rate": 2.171320357287626e-06, "loss": 0.0175, "step": 145690 }, { "epoch": 2.91155429439271, "grad_norm": 0.10611971467733383, "learning_rate": 2.1683090573072275e-06, "loss": 0.0189, "step": 145700 }, { "epoch": 2.9117541265337117, "grad_norm": 0.10819453001022339, "learning_rate": 2.1652998006088e-06, "loss": 0.0215, "step": 145710 }, { "epoch": 2.9119539586747134, "grad_norm": 0.15254418551921844, "learning_rate": 2.162292587320891e-06, "loss": 0.0258, "step": 145720 }, { "epoch": 2.9121537908157147, "grad_norm": 0.12203679978847504, "learning_rate": 2.1592874175719592e-06, "loss": 0.0268, "step": 145730 }, { "epoch": 2.9123536229567164, "grad_norm": 0.10201635211706161, "learning_rate": 2.1562842914903845e-06, "loss": 0.0252, "step": 145740 }, { "epoch": 2.9125534550977177, "grad_norm": 0.10572712868452072, "learning_rate": 2.1532832092044595e-06, "loss": 0.0297, "step": 145750 }, { "epoch": 2.9127532872387194, "grad_norm": 0.17631320655345917, "learning_rate": 2.1502841708423704e-06, "loss": 0.0237, "step": 145760 }, { "epoch": 2.912953119379721, "grad_norm": 0.09371522068977356, "learning_rate": 2.1472871765322313e-06, "loss": 0.0236, "step": 145770 }, { "epoch": 2.913152951520723, "grad_norm": 0.12479248642921448, "learning_rate": 2.144292226402095e-06, "loss": 0.0229, "step": 145780 }, { "epoch": 2.913352783661724, "grad_norm": 0.12420447170734406, "learning_rate": 2.1412993205798705e-06, "loss": 0.0216, "step": 145790 }, { "epoch": 2.913552615802726, "grad_norm": 0.1267423927783966, "learning_rate": 2.138308459193417e-06, "loss": 0.0258, "step": 145800 }, { "epoch": 2.913752447943727, "grad_norm": 0.13775114715099335, "learning_rate": 2.1353196423705037e-06, "loss": 0.0299, "step": 145810 }, { "epoch": 2.913952280084729, "grad_norm": 0.09750384092330933, "learning_rate": 2.132332870238807e-06, "loss": 0.0252, "step": 145820 }, { "epoch": 2.9141521122257306, "grad_norm": 0.09368223696947098, "learning_rate": 2.1293481429259133e-06, "loss": 0.025, "step": 145830 }, { "epoch": 2.914351944366732, "grad_norm": 0.11467588692903519, "learning_rate": 2.1263654605593265e-06, "loss": 0.0274, "step": 145840 }, { "epoch": 2.9145517765077336, "grad_norm": 0.11509336531162262, "learning_rate": 2.123384823266461e-06, "loss": 0.0244, "step": 145850 }, { "epoch": 2.914751608648735, "grad_norm": 0.1554252803325653, "learning_rate": 2.1204062311746373e-06, "loss": 0.0239, "step": 145860 }, { "epoch": 2.9149514407897366, "grad_norm": 0.08789003640413284, "learning_rate": 2.117429684411104e-06, "loss": 0.0178, "step": 145870 }, { "epoch": 2.9151512729307383, "grad_norm": 0.09623385220766068, "learning_rate": 2.114455183103026e-06, "loss": 0.0172, "step": 145880 }, { "epoch": 2.91535110507174, "grad_norm": 0.08983851969242096, "learning_rate": 2.1114827273774397e-06, "loss": 0.0288, "step": 145890 }, { "epoch": 2.9155509372127413, "grad_norm": 0.09271547198295593, "learning_rate": 2.108512317361333e-06, "loss": 0.0285, "step": 145900 }, { "epoch": 2.915750769353743, "grad_norm": 0.09751112014055252, "learning_rate": 2.105543953181616e-06, "loss": 0.0218, "step": 145910 }, { "epoch": 2.9159506014947443, "grad_norm": 0.10271508246660233, "learning_rate": 2.102577634965064e-06, "loss": 0.0239, "step": 145920 }, { "epoch": 2.916150433635746, "grad_norm": 0.08426477015018463, "learning_rate": 2.0996133628384095e-06, "loss": 0.0346, "step": 145930 }, { "epoch": 2.9163502657767477, "grad_norm": 0.11271875351667404, "learning_rate": 2.096651136928274e-06, "loss": 0.0263, "step": 145940 }, { "epoch": 2.916550097917749, "grad_norm": 0.16100570559501648, "learning_rate": 2.093690957361205e-06, "loss": 0.033, "step": 145950 }, { "epoch": 2.9167499300587507, "grad_norm": 0.13651521503925323, "learning_rate": 2.090732824263647e-06, "loss": 0.0202, "step": 145960 }, { "epoch": 2.916949762199752, "grad_norm": 0.11743409186601639, "learning_rate": 2.0877767377619705e-06, "loss": 0.0251, "step": 145970 }, { "epoch": 2.9171495943407537, "grad_norm": 0.09430316835641861, "learning_rate": 2.0848226979824638e-06, "loss": 0.0209, "step": 145980 }, { "epoch": 2.9173494264817554, "grad_norm": 0.15512432157993317, "learning_rate": 2.0818707050512974e-06, "loss": 0.0227, "step": 145990 }, { "epoch": 2.917549258622757, "grad_norm": 0.09925093501806259, "learning_rate": 2.078920759094588e-06, "loss": 0.0289, "step": 146000 }, { "epoch": 2.9177490907637584, "grad_norm": 0.08047594130039215, "learning_rate": 2.0759728602383567e-06, "loss": 0.0244, "step": 146010 }, { "epoch": 2.91794892290476, "grad_norm": 0.15143363177776337, "learning_rate": 2.073027008608519e-06, "loss": 0.0233, "step": 146020 }, { "epoch": 2.9181487550457614, "grad_norm": 0.1884559690952301, "learning_rate": 2.0700832043309247e-06, "loss": 0.0218, "step": 146030 }, { "epoch": 2.918348587186763, "grad_norm": 0.10734576731920242, "learning_rate": 2.0671414475313177e-06, "loss": 0.0239, "step": 146040 }, { "epoch": 2.918548419327765, "grad_norm": 0.11493837088346481, "learning_rate": 2.0642017383353863e-06, "loss": 0.0218, "step": 146050 }, { "epoch": 2.918748251468766, "grad_norm": 0.11628089845180511, "learning_rate": 2.061264076868691e-06, "loss": 0.0245, "step": 146060 }, { "epoch": 2.918948083609768, "grad_norm": 0.14095556735992432, "learning_rate": 2.058328463256726e-06, "loss": 0.0183, "step": 146070 }, { "epoch": 2.919147915750769, "grad_norm": 0.12169288843870163, "learning_rate": 2.055394897624896e-06, "loss": 0.0229, "step": 146080 }, { "epoch": 2.919347747891771, "grad_norm": 0.1618172526359558, "learning_rate": 2.052463380098524e-06, "loss": 0.0339, "step": 146090 }, { "epoch": 2.9195475800327726, "grad_norm": 0.13390478491783142, "learning_rate": 2.049533910802831e-06, "loss": 0.0245, "step": 146100 }, { "epoch": 2.919747412173774, "grad_norm": 0.14181894063949585, "learning_rate": 2.0466064898629624e-06, "loss": 0.0303, "step": 146110 }, { "epoch": 2.9199472443147756, "grad_norm": 0.10833822935819626, "learning_rate": 2.043681117403973e-06, "loss": 0.0241, "step": 146120 }, { "epoch": 2.9201470764557773, "grad_norm": 0.09112269431352615, "learning_rate": 2.040757793550818e-06, "loss": 0.0219, "step": 146130 }, { "epoch": 2.9203469085967786, "grad_norm": 0.12441570311784744, "learning_rate": 2.037836518428393e-06, "loss": 0.0211, "step": 146140 }, { "epoch": 2.9205467407377803, "grad_norm": 0.13924139738082886, "learning_rate": 2.034917292161492e-06, "loss": 0.0177, "step": 146150 }, { "epoch": 2.920746572878782, "grad_norm": 0.1339653730392456, "learning_rate": 2.032000114874799e-06, "loss": 0.0239, "step": 146160 }, { "epoch": 2.9209464050197833, "grad_norm": 0.14135050773620605, "learning_rate": 2.029084986692936e-06, "loss": 0.026, "step": 146170 }, { "epoch": 2.921146237160785, "grad_norm": 0.16331368684768677, "learning_rate": 2.0261719077404483e-06, "loss": 0.0317, "step": 146180 }, { "epoch": 2.9213460693017863, "grad_norm": 0.12764869630336761, "learning_rate": 2.023260878141764e-06, "loss": 0.0256, "step": 146190 }, { "epoch": 2.921545901442788, "grad_norm": 0.12434054911136627, "learning_rate": 2.0203518980212334e-06, "loss": 0.0223, "step": 146200 }, { "epoch": 2.9217457335837898, "grad_norm": 0.13271570205688477, "learning_rate": 2.0174449675031347e-06, "loss": 0.0208, "step": 146210 }, { "epoch": 2.921945565724791, "grad_norm": 0.12132707238197327, "learning_rate": 2.014540086711636e-06, "loss": 0.0212, "step": 146220 }, { "epoch": 2.9221453978657927, "grad_norm": 0.14704324305057526, "learning_rate": 2.011637255770832e-06, "loss": 0.0228, "step": 146230 }, { "epoch": 2.922345230006794, "grad_norm": 0.09266616404056549, "learning_rate": 2.0087364748047286e-06, "loss": 0.0235, "step": 146240 }, { "epoch": 2.9225450621477957, "grad_norm": 0.12653571367263794, "learning_rate": 2.0058377439372446e-06, "loss": 0.0277, "step": 146250 }, { "epoch": 2.9227448942887975, "grad_norm": 0.11394231021404266, "learning_rate": 2.0029410632921907e-06, "loss": 0.0198, "step": 146260 }, { "epoch": 2.922944726429799, "grad_norm": 0.11006394773721695, "learning_rate": 2.00004643299333e-06, "loss": 0.0198, "step": 146270 }, { "epoch": 2.9231445585708005, "grad_norm": 0.11246161162853241, "learning_rate": 1.997153853164313e-06, "loss": 0.0264, "step": 146280 }, { "epoch": 2.923344390711802, "grad_norm": 0.14408466219902039, "learning_rate": 1.994263323928697e-06, "loss": 0.0313, "step": 146290 }, { "epoch": 2.9235442228528035, "grad_norm": 0.13159437477588654, "learning_rate": 1.9913748454099544e-06, "loss": 0.0228, "step": 146300 }, { "epoch": 2.923744054993805, "grad_norm": 0.1788794994354248, "learning_rate": 1.988488417731488e-06, "loss": 0.024, "step": 146310 }, { "epoch": 2.923943887134807, "grad_norm": 0.11679920554161072, "learning_rate": 1.985604041016598e-06, "loss": 0.0225, "step": 146320 }, { "epoch": 2.924143719275808, "grad_norm": 0.12188543379306793, "learning_rate": 1.982721715388497e-06, "loss": 0.0323, "step": 146330 }, { "epoch": 2.92434355141681, "grad_norm": 0.13332964479923248, "learning_rate": 1.9798414409703093e-06, "loss": 0.0248, "step": 146340 }, { "epoch": 2.924543383557811, "grad_norm": 0.13014477491378784, "learning_rate": 1.976963217885086e-06, "loss": 0.0255, "step": 146350 }, { "epoch": 2.924743215698813, "grad_norm": 0.1217435970902443, "learning_rate": 1.974087046255768e-06, "loss": 0.0225, "step": 146360 }, { "epoch": 2.9249430478398146, "grad_norm": 0.12352567911148071, "learning_rate": 1.9712129262052293e-06, "loss": 0.0251, "step": 146370 }, { "epoch": 2.9251428799808163, "grad_norm": 0.13090161979198456, "learning_rate": 1.9683408578562433e-06, "loss": 0.0219, "step": 146380 }, { "epoch": 2.9253427121218176, "grad_norm": 0.1650746762752533, "learning_rate": 1.965470841331496e-06, "loss": 0.0226, "step": 146390 }, { "epoch": 2.9255425442628193, "grad_norm": 0.13683831691741943, "learning_rate": 1.9626028767535886e-06, "loss": 0.0214, "step": 146400 }, { "epoch": 2.9257423764038206, "grad_norm": 0.09287083894014359, "learning_rate": 1.9597369642450457e-06, "loss": 0.0181, "step": 146410 }, { "epoch": 2.9259422085448223, "grad_norm": 0.10884741693735123, "learning_rate": 1.956873103928286e-06, "loss": 0.0223, "step": 146420 }, { "epoch": 2.926142040685824, "grad_norm": 0.0887727290391922, "learning_rate": 1.954011295925645e-06, "loss": 0.0216, "step": 146430 }, { "epoch": 2.9263418728268253, "grad_norm": 0.07890094816684723, "learning_rate": 1.951151540359375e-06, "loss": 0.0204, "step": 146440 }, { "epoch": 2.926541704967827, "grad_norm": 0.21696996688842773, "learning_rate": 1.94829383735165e-06, "loss": 0.0227, "step": 146450 }, { "epoch": 2.9267415371088283, "grad_norm": 0.1504908949136734, "learning_rate": 1.9454381870245396e-06, "loss": 0.0216, "step": 146460 }, { "epoch": 2.92694136924983, "grad_norm": 0.07467423379421234, "learning_rate": 1.942584589500024e-06, "loss": 0.0254, "step": 146470 }, { "epoch": 2.927141201390832, "grad_norm": 0.07757315039634705, "learning_rate": 1.939733044900016e-06, "loss": 0.03, "step": 146480 }, { "epoch": 2.9273410335318335, "grad_norm": 0.13104121387004852, "learning_rate": 1.936883553346319e-06, "loss": 0.0236, "step": 146490 }, { "epoch": 2.927540865672835, "grad_norm": 0.11606571078300476, "learning_rate": 1.9340361149606632e-06, "loss": 0.0242, "step": 146500 }, { "epoch": 2.9277406978138365, "grad_norm": 0.08500448614358902, "learning_rate": 1.9311907298646847e-06, "loss": 0.0281, "step": 146510 }, { "epoch": 2.927940529954838, "grad_norm": 0.14263448119163513, "learning_rate": 1.928347398179936e-06, "loss": 0.0273, "step": 146520 }, { "epoch": 2.9281403620958395, "grad_norm": 0.12229304015636444, "learning_rate": 1.92550612002787e-06, "loss": 0.0244, "step": 146530 }, { "epoch": 2.928340194236841, "grad_norm": 0.13670118153095245, "learning_rate": 1.9226668955298675e-06, "loss": 0.0346, "step": 146540 }, { "epoch": 2.9285400263778425, "grad_norm": 0.11951036006212234, "learning_rate": 1.9198297248072263e-06, "loss": 0.0193, "step": 146550 }, { "epoch": 2.928739858518844, "grad_norm": 0.12996044754981995, "learning_rate": 1.916994607981121e-06, "loss": 0.0306, "step": 146560 }, { "epoch": 2.9289396906598455, "grad_norm": 0.09741540998220444, "learning_rate": 1.914161545172677e-06, "loss": 0.0218, "step": 146570 }, { "epoch": 2.929139522800847, "grad_norm": 0.10468007624149323, "learning_rate": 1.9113305365029144e-06, "loss": 0.0227, "step": 146580 }, { "epoch": 2.929339354941849, "grad_norm": 0.07988116145133972, "learning_rate": 1.90850158209277e-06, "loss": 0.0224, "step": 146590 }, { "epoch": 2.9295391870828507, "grad_norm": 0.11901137232780457, "learning_rate": 1.9056746820630911e-06, "loss": 0.0291, "step": 146600 }, { "epoch": 2.929739019223852, "grad_norm": 0.13214750587940216, "learning_rate": 1.9028498365346425e-06, "loss": 0.0235, "step": 146610 }, { "epoch": 2.9299388513648537, "grad_norm": 0.19363485276699066, "learning_rate": 1.9000270456280834e-06, "loss": 0.0237, "step": 146620 }, { "epoch": 2.930138683505855, "grad_norm": 0.12162105739116669, "learning_rate": 1.8972063094640114e-06, "loss": 0.0243, "step": 146630 }, { "epoch": 2.9303385156468567, "grad_norm": 0.11811211705207825, "learning_rate": 1.8943876281629192e-06, "loss": 0.0276, "step": 146640 }, { "epoch": 2.9305383477878584, "grad_norm": 0.07689971476793289, "learning_rate": 1.891571001845216e-06, "loss": 0.0231, "step": 146650 }, { "epoch": 2.9307381799288597, "grad_norm": 0.12903523445129395, "learning_rate": 1.8887564306312222e-06, "loss": 0.0185, "step": 146660 }, { "epoch": 2.9309380120698614, "grad_norm": 0.1613130122423172, "learning_rate": 1.8859439146411585e-06, "loss": 0.019, "step": 146670 }, { "epoch": 2.9311378442108627, "grad_norm": 0.09438107162714005, "learning_rate": 1.883133453995195e-06, "loss": 0.0271, "step": 146680 }, { "epoch": 2.9313376763518644, "grad_norm": 0.16247187554836273, "learning_rate": 1.880325048813375e-06, "loss": 0.0271, "step": 146690 }, { "epoch": 2.931537508492866, "grad_norm": 0.12797121703624725, "learning_rate": 1.877518699215669e-06, "loss": 0.0248, "step": 146700 }, { "epoch": 2.931737340633868, "grad_norm": 0.17215226590633392, "learning_rate": 1.874714405321959e-06, "loss": 0.0204, "step": 146710 }, { "epoch": 2.931937172774869, "grad_norm": 0.14254415035247803, "learning_rate": 1.8719121672520378e-06, "loss": 0.0203, "step": 146720 }, { "epoch": 2.932137004915871, "grad_norm": 0.14515084028244019, "learning_rate": 1.8691119851256156e-06, "loss": 0.0201, "step": 146730 }, { "epoch": 2.932336837056872, "grad_norm": 0.12071534246206284, "learning_rate": 1.866313859062313e-06, "loss": 0.0192, "step": 146740 }, { "epoch": 2.932536669197874, "grad_norm": 0.1030021533370018, "learning_rate": 1.8635177891816624e-06, "loss": 0.0219, "step": 146750 }, { "epoch": 2.9327365013388755, "grad_norm": 0.07887829095125198, "learning_rate": 1.8607237756030904e-06, "loss": 0.0263, "step": 146760 }, { "epoch": 2.932936333479877, "grad_norm": 0.10093183815479279, "learning_rate": 1.8579318184459737e-06, "loss": 0.0212, "step": 146770 }, { "epoch": 2.9331361656208785, "grad_norm": 0.10713531821966171, "learning_rate": 1.8551419178295726e-06, "loss": 0.0227, "step": 146780 }, { "epoch": 2.93333599776188, "grad_norm": 0.10912003368139267, "learning_rate": 1.8523540738730582e-06, "loss": 0.0206, "step": 146790 }, { "epoch": 2.9335358299028815, "grad_norm": 0.11228426545858383, "learning_rate": 1.8495682866955189e-06, "loss": 0.0296, "step": 146800 }, { "epoch": 2.9337356620438833, "grad_norm": 0.10691242665052414, "learning_rate": 1.8467845564159814e-06, "loss": 0.0333, "step": 146810 }, { "epoch": 2.933935494184885, "grad_norm": 0.11338485032320023, "learning_rate": 1.8440028831533452e-06, "loss": 0.0231, "step": 146820 }, { "epoch": 2.9341353263258863, "grad_norm": 0.09578040987253189, "learning_rate": 1.8412232670264372e-06, "loss": 0.0219, "step": 146830 }, { "epoch": 2.934335158466888, "grad_norm": 0.11859842389822006, "learning_rate": 1.8384457081539909e-06, "loss": 0.0306, "step": 146840 }, { "epoch": 2.9345349906078892, "grad_norm": 0.0923205092549324, "learning_rate": 1.8356702066546828e-06, "loss": 0.0182, "step": 146850 }, { "epoch": 2.934734822748891, "grad_norm": 0.12729275226593018, "learning_rate": 1.8328967626470628e-06, "loss": 0.0224, "step": 146860 }, { "epoch": 2.9349346548898927, "grad_norm": 0.12305516004562378, "learning_rate": 1.8301253762495973e-06, "loss": 0.0196, "step": 146870 }, { "epoch": 2.935134487030894, "grad_norm": 0.1309889405965805, "learning_rate": 1.8273560475806917e-06, "loss": 0.0227, "step": 146880 }, { "epoch": 2.9353343191718957, "grad_norm": 0.14753536880016327, "learning_rate": 1.82458877675864e-06, "loss": 0.0287, "step": 146890 }, { "epoch": 2.935534151312897, "grad_norm": 0.0920238047838211, "learning_rate": 1.8218235639016533e-06, "loss": 0.0228, "step": 146900 }, { "epoch": 2.9357339834538987, "grad_norm": 0.1044812873005867, "learning_rate": 1.819060409127854e-06, "loss": 0.0233, "step": 146910 }, { "epoch": 2.9359338155949004, "grad_norm": 0.10696998238563538, "learning_rate": 1.8162993125552918e-06, "loss": 0.0304, "step": 146920 }, { "epoch": 2.9361336477359017, "grad_norm": 0.12771326303482056, "learning_rate": 1.8135402743018947e-06, "loss": 0.0205, "step": 146930 }, { "epoch": 2.9363334798769034, "grad_norm": 0.12081944197416306, "learning_rate": 1.8107832944855462e-06, "loss": 0.0275, "step": 146940 }, { "epoch": 2.9365333120179047, "grad_norm": 0.13209271430969238, "learning_rate": 1.8080283732240078e-06, "loss": 0.0222, "step": 146950 }, { "epoch": 2.9367331441589064, "grad_norm": 0.1176019087433815, "learning_rate": 1.8052755106349628e-06, "loss": 0.0208, "step": 146960 }, { "epoch": 2.936932976299908, "grad_norm": 0.09225594252347946, "learning_rate": 1.8025247068360117e-06, "loss": 0.018, "step": 146970 }, { "epoch": 2.93713280844091, "grad_norm": 0.1137939915060997, "learning_rate": 1.7997759619446607e-06, "loss": 0.0243, "step": 146980 }, { "epoch": 2.937332640581911, "grad_norm": 0.09399078041315079, "learning_rate": 1.797029276078338e-06, "loss": 0.0237, "step": 146990 }, { "epoch": 2.937532472722913, "grad_norm": 0.11434612423181534, "learning_rate": 1.794284649354372e-06, "loss": 0.0283, "step": 147000 }, { "epoch": 2.937732304863914, "grad_norm": 0.21145862340927124, "learning_rate": 1.7915420818900076e-06, "loss": 0.0222, "step": 147010 }, { "epoch": 2.937932137004916, "grad_norm": 0.11672452092170715, "learning_rate": 1.7888015738024066e-06, "loss": 0.0231, "step": 147020 }, { "epoch": 2.9381319691459176, "grad_norm": 0.08612487465143204, "learning_rate": 1.7860631252086313e-06, "loss": 0.0177, "step": 147030 }, { "epoch": 2.938331801286919, "grad_norm": 0.15268844366073608, "learning_rate": 1.7833267362256711e-06, "loss": 0.0182, "step": 147040 }, { "epoch": 2.9385316334279206, "grad_norm": 0.11870303004980087, "learning_rate": 1.7805924069704216e-06, "loss": 0.0273, "step": 147050 }, { "epoch": 2.938731465568922, "grad_norm": 0.15363217890262604, "learning_rate": 1.7778601375596727e-06, "loss": 0.0228, "step": 147060 }, { "epoch": 2.9389312977099236, "grad_norm": 0.10720318555831909, "learning_rate": 1.7751299281101474e-06, "loss": 0.0258, "step": 147070 }, { "epoch": 2.9391311298509253, "grad_norm": 0.1325346976518631, "learning_rate": 1.7724017787384972e-06, "loss": 0.0336, "step": 147080 }, { "epoch": 2.939330961991927, "grad_norm": 0.0983474999666214, "learning_rate": 1.7696756895612343e-06, "loss": 0.0173, "step": 147090 }, { "epoch": 2.9395307941329283, "grad_norm": 0.08281169831752777, "learning_rate": 1.7669516606948266e-06, "loss": 0.0222, "step": 147100 }, { "epoch": 2.93973062627393, "grad_norm": 0.11452226340770721, "learning_rate": 1.7642296922556313e-06, "loss": 0.028, "step": 147110 }, { "epoch": 2.9399304584149313, "grad_norm": 0.17079761624336243, "learning_rate": 1.761509784359938e-06, "loss": 0.0204, "step": 147120 }, { "epoch": 2.940130290555933, "grad_norm": 0.07513907551765442, "learning_rate": 1.758791937123927e-06, "loss": 0.0252, "step": 147130 }, { "epoch": 2.9403301226969347, "grad_norm": 0.12834645807743073, "learning_rate": 1.7560761506637046e-06, "loss": 0.0226, "step": 147140 }, { "epoch": 2.940529954837936, "grad_norm": 0.11642565578222275, "learning_rate": 1.753362425095284e-06, "loss": 0.0292, "step": 147150 }, { "epoch": 2.9407297869789377, "grad_norm": 0.22306963801383972, "learning_rate": 1.7506507605345834e-06, "loss": 0.0184, "step": 147160 }, { "epoch": 2.940929619119939, "grad_norm": 0.12109046429395676, "learning_rate": 1.7479411570974492e-06, "loss": 0.0238, "step": 147170 }, { "epoch": 2.9411294512609407, "grad_norm": 0.1286730319261551, "learning_rate": 1.7452336148996273e-06, "loss": 0.0134, "step": 147180 }, { "epoch": 2.9413292834019424, "grad_norm": 0.1392606794834137, "learning_rate": 1.7425281340567756e-06, "loss": 0.0156, "step": 147190 }, { "epoch": 2.941529115542944, "grad_norm": 0.1601717472076416, "learning_rate": 1.7398247146844682e-06, "loss": 0.0263, "step": 147200 }, { "epoch": 2.9417289476839454, "grad_norm": 0.09583652019500732, "learning_rate": 1.737123356898196e-06, "loss": 0.0192, "step": 147210 }, { "epoch": 2.941928779824947, "grad_norm": 0.14329442381858826, "learning_rate": 1.7344240608133554e-06, "loss": 0.0305, "step": 147220 }, { "epoch": 2.9421286119659484, "grad_norm": 0.10891502350568771, "learning_rate": 1.731726826545249e-06, "loss": 0.0307, "step": 147230 }, { "epoch": 2.94232844410695, "grad_norm": 0.12353407591581345, "learning_rate": 1.7290316542091012e-06, "loss": 0.0317, "step": 147240 }, { "epoch": 2.942528276247952, "grad_norm": 0.09677174687385559, "learning_rate": 1.726338543920042e-06, "loss": 0.0171, "step": 147250 }, { "epoch": 2.942728108388953, "grad_norm": 0.08601976186037064, "learning_rate": 1.7236474957931182e-06, "loss": 0.0215, "step": 147260 }, { "epoch": 2.942927940529955, "grad_norm": 0.12845812737941742, "learning_rate": 1.7209585099432934e-06, "loss": 0.0235, "step": 147270 }, { "epoch": 2.943127772670956, "grad_norm": 0.14825844764709473, "learning_rate": 1.7182715864854204e-06, "loss": 0.0248, "step": 147280 }, { "epoch": 2.943327604811958, "grad_norm": 0.12655867636203766, "learning_rate": 1.715586725534296e-06, "loss": 0.0229, "step": 147290 }, { "epoch": 2.9435274369529596, "grad_norm": 0.09054803848266602, "learning_rate": 1.7129039272046011e-06, "loss": 0.0185, "step": 147300 }, { "epoch": 2.9437272690939613, "grad_norm": 0.11638520658016205, "learning_rate": 1.710223191610949e-06, "loss": 0.0241, "step": 147310 }, { "epoch": 2.9439271012349626, "grad_norm": 0.1257942020893097, "learning_rate": 1.707544518867854e-06, "loss": 0.0287, "step": 147320 }, { "epoch": 2.9441269333759643, "grad_norm": 0.1895235776901245, "learning_rate": 1.7048679090897357e-06, "loss": 0.0328, "step": 147330 }, { "epoch": 2.9443267655169656, "grad_norm": 0.15404963493347168, "learning_rate": 1.7021933623909303e-06, "loss": 0.0152, "step": 147340 }, { "epoch": 2.9445265976579673, "grad_norm": 0.11662460118532181, "learning_rate": 1.699520878885713e-06, "loss": 0.019, "step": 147350 }, { "epoch": 2.944726429798969, "grad_norm": 0.10788512229919434, "learning_rate": 1.6968504586882262e-06, "loss": 0.0209, "step": 147360 }, { "epoch": 2.9449262619399703, "grad_norm": 0.11364684253931046, "learning_rate": 1.6941821019125558e-06, "loss": 0.0172, "step": 147370 }, { "epoch": 2.945126094080972, "grad_norm": 0.1118423342704773, "learning_rate": 1.691515808672678e-06, "loss": 0.0202, "step": 147380 }, { "epoch": 2.9453259262219733, "grad_norm": 0.10801014304161072, "learning_rate": 1.6888515790825066e-06, "loss": 0.0275, "step": 147390 }, { "epoch": 2.945525758362975, "grad_norm": 0.14829470217227936, "learning_rate": 1.6861894132558398e-06, "loss": 0.0258, "step": 147400 }, { "epoch": 2.9457255905039768, "grad_norm": 0.10658735036849976, "learning_rate": 1.6835293113064089e-06, "loss": 0.0294, "step": 147410 }, { "epoch": 2.9459254226449785, "grad_norm": 0.12513041496276855, "learning_rate": 1.6808712733478503e-06, "loss": 0.0317, "step": 147420 }, { "epoch": 2.9461252547859798, "grad_norm": 0.08908648043870926, "learning_rate": 1.6782152994936961e-06, "loss": 0.0234, "step": 147430 }, { "epoch": 2.9463250869269815, "grad_norm": 0.11987978219985962, "learning_rate": 1.675561389857422e-06, "loss": 0.0194, "step": 147440 }, { "epoch": 2.9465249190679828, "grad_norm": 0.13572882115840912, "learning_rate": 1.6729095445523923e-06, "loss": 0.0303, "step": 147450 }, { "epoch": 2.9467247512089845, "grad_norm": 0.2165777087211609, "learning_rate": 1.6702597636918838e-06, "loss": 0.0283, "step": 147460 }, { "epoch": 2.946924583349986, "grad_norm": 0.15208545327186584, "learning_rate": 1.6676120473890832e-06, "loss": 0.0255, "step": 147470 }, { "epoch": 2.9471244154909875, "grad_norm": 0.15549129247665405, "learning_rate": 1.6649663957571226e-06, "loss": 0.0231, "step": 147480 }, { "epoch": 2.947324247631989, "grad_norm": 0.11458875983953476, "learning_rate": 1.6623228089089948e-06, "loss": 0.0187, "step": 147490 }, { "epoch": 2.9475240797729905, "grad_norm": 0.13319270312786102, "learning_rate": 1.6596812869576428e-06, "loss": 0.0267, "step": 147500 }, { "epoch": 2.947723911913992, "grad_norm": 0.13911467790603638, "learning_rate": 1.657041830015893e-06, "loss": 0.0251, "step": 147510 }, { "epoch": 2.947923744054994, "grad_norm": 0.0778343603014946, "learning_rate": 1.6544044381965163e-06, "loss": 0.0234, "step": 147520 }, { "epoch": 2.9481235761959956, "grad_norm": 0.13406288623809814, "learning_rate": 1.6517691116121614e-06, "loss": 0.0232, "step": 147530 }, { "epoch": 2.948323408336997, "grad_norm": 0.1017441526055336, "learning_rate": 1.6491358503754163e-06, "loss": 0.0196, "step": 147540 }, { "epoch": 2.9485232404779986, "grad_norm": 0.1204899400472641, "learning_rate": 1.6465046545987684e-06, "loss": 0.0242, "step": 147550 }, { "epoch": 2.948723072619, "grad_norm": 0.09788887202739716, "learning_rate": 1.6438755243946002e-06, "loss": 0.0314, "step": 147560 }, { "epoch": 2.9489229047600016, "grad_norm": 0.0947485938668251, "learning_rate": 1.6412484598752497e-06, "loss": 0.0263, "step": 147570 }, { "epoch": 2.9491227369010033, "grad_norm": 0.0917803943157196, "learning_rate": 1.6386234611529273e-06, "loss": 0.0168, "step": 147580 }, { "epoch": 2.9493225690420046, "grad_norm": 0.12893761694431305, "learning_rate": 1.6360005283397594e-06, "loss": 0.0286, "step": 147590 }, { "epoch": 2.9495224011830063, "grad_norm": 0.16005876660346985, "learning_rate": 1.633379661547807e-06, "loss": 0.0203, "step": 147600 }, { "epoch": 2.9497222333240076, "grad_norm": 0.1458316445350647, "learning_rate": 1.6307608608890134e-06, "loss": 0.021, "step": 147610 }, { "epoch": 2.9499220654650093, "grad_norm": 0.10953941941261292, "learning_rate": 1.6281441264752728e-06, "loss": 0.03, "step": 147620 }, { "epoch": 2.950121897606011, "grad_norm": 0.09821062535047531, "learning_rate": 1.6255294584183456e-06, "loss": 0.0249, "step": 147630 }, { "epoch": 2.9503217297470123, "grad_norm": 0.14018550515174866, "learning_rate": 1.6229168568299313e-06, "loss": 0.0208, "step": 147640 }, { "epoch": 2.950521561888014, "grad_norm": 0.16066251695156097, "learning_rate": 1.6203063218216407e-06, "loss": 0.0255, "step": 147650 }, { "epoch": 2.950721394029016, "grad_norm": 0.12325994670391083, "learning_rate": 1.6176978535049847e-06, "loss": 0.024, "step": 147660 }, { "epoch": 2.950921226170017, "grad_norm": 0.09783048927783966, "learning_rate": 1.6150914519913962e-06, "loss": 0.0224, "step": 147670 }, { "epoch": 2.951121058311019, "grad_norm": 0.09890920668840408, "learning_rate": 1.6124871173922196e-06, "loss": 0.027, "step": 147680 }, { "epoch": 2.9513208904520205, "grad_norm": 0.12950929999351501, "learning_rate": 1.6098848498187046e-06, "loss": 0.021, "step": 147690 }, { "epoch": 2.951520722593022, "grad_norm": 0.13617068529129028, "learning_rate": 1.6072846493820014e-06, "loss": 0.0238, "step": 147700 }, { "epoch": 2.9517205547340235, "grad_norm": 0.10917828232049942, "learning_rate": 1.6046865161932045e-06, "loss": 0.0238, "step": 147710 }, { "epoch": 2.951920386875025, "grad_norm": 0.116531141102314, "learning_rate": 1.6020904503632973e-06, "loss": 0.0266, "step": 147720 }, { "epoch": 2.9521202190160265, "grad_norm": 0.13145601749420166, "learning_rate": 1.5994964520031742e-06, "loss": 0.0211, "step": 147730 }, { "epoch": 2.9523200511570282, "grad_norm": 0.13637278974056244, "learning_rate": 1.5969045212236412e-06, "loss": 0.0183, "step": 147740 }, { "epoch": 2.9525198832980295, "grad_norm": 0.0904986709356308, "learning_rate": 1.5943146581354428e-06, "loss": 0.0246, "step": 147750 }, { "epoch": 2.9527197154390312, "grad_norm": 0.10170965641736984, "learning_rate": 1.5917268628491854e-06, "loss": 0.0209, "step": 147760 }, { "epoch": 2.9529195475800325, "grad_norm": 0.09748254716396332, "learning_rate": 1.5891411354754305e-06, "loss": 0.0227, "step": 147770 }, { "epoch": 2.953119379721034, "grad_norm": 0.15098980069160461, "learning_rate": 1.5865574761246338e-06, "loss": 0.028, "step": 147780 }, { "epoch": 2.953319211862036, "grad_norm": 0.0941067785024643, "learning_rate": 1.583975884907163e-06, "loss": 0.0181, "step": 147790 }, { "epoch": 2.9535190440030377, "grad_norm": 0.15407879650592804, "learning_rate": 1.5813963619333016e-06, "loss": 0.0255, "step": 147800 }, { "epoch": 2.953718876144039, "grad_norm": 0.12723097205162048, "learning_rate": 1.5788189073132342e-06, "loss": 0.0315, "step": 147810 }, { "epoch": 2.9539187082850407, "grad_norm": 0.12927411496639252, "learning_rate": 1.5762435211570836e-06, "loss": 0.0187, "step": 147820 }, { "epoch": 2.954118540426042, "grad_norm": 0.10532251000404358, "learning_rate": 1.5736702035748341e-06, "loss": 0.0227, "step": 147830 }, { "epoch": 2.9543183725670437, "grad_norm": 0.11250341683626175, "learning_rate": 1.571098954676442e-06, "loss": 0.0216, "step": 147840 }, { "epoch": 2.9545182047080454, "grad_norm": 0.09634029865264893, "learning_rate": 1.5685297745717363e-06, "loss": 0.0228, "step": 147850 }, { "epoch": 2.9547180368490467, "grad_norm": 0.16564953327178955, "learning_rate": 1.5659626633704682e-06, "loss": 0.0313, "step": 147860 }, { "epoch": 2.9549178689900484, "grad_norm": 0.0813855305314064, "learning_rate": 1.5633976211822943e-06, "loss": 0.0178, "step": 147870 }, { "epoch": 2.9551177011310497, "grad_norm": 0.12248361110687256, "learning_rate": 1.5608346481167934e-06, "loss": 0.0255, "step": 147880 }, { "epoch": 2.9553175332720514, "grad_norm": 0.13391761481761932, "learning_rate": 1.5582737442834505e-06, "loss": 0.0335, "step": 147890 }, { "epoch": 2.955517365413053, "grad_norm": 0.10395753383636475, "learning_rate": 1.5557149097916668e-06, "loss": 0.0241, "step": 147900 }, { "epoch": 2.955717197554055, "grad_norm": 0.15791645646095276, "learning_rate": 1.5531581447507437e-06, "loss": 0.0183, "step": 147910 }, { "epoch": 2.955917029695056, "grad_norm": 0.09393437206745148, "learning_rate": 1.550603449269905e-06, "loss": 0.0223, "step": 147920 }, { "epoch": 2.956116861836058, "grad_norm": 0.134514719247818, "learning_rate": 1.5480508234582858e-06, "loss": 0.0213, "step": 147930 }, { "epoch": 2.956316693977059, "grad_norm": 0.08840730041265488, "learning_rate": 1.5455002674249207e-06, "loss": 0.0222, "step": 147940 }, { "epoch": 2.956516526118061, "grad_norm": 0.11283545941114426, "learning_rate": 1.5429517812787786e-06, "loss": 0.0199, "step": 147950 }, { "epoch": 2.9567163582590625, "grad_norm": 0.13539937138557434, "learning_rate": 1.5404053651287164e-06, "loss": 0.0177, "step": 147960 }, { "epoch": 2.956916190400064, "grad_norm": 0.13143743574619293, "learning_rate": 1.537861019083503e-06, "loss": 0.0188, "step": 147970 }, { "epoch": 2.9571160225410655, "grad_norm": 0.10434878617525101, "learning_rate": 1.5353187432518511e-06, "loss": 0.024, "step": 147980 }, { "epoch": 2.957315854682067, "grad_norm": 0.15560446679592133, "learning_rate": 1.5327785377423464e-06, "loss": 0.0269, "step": 147990 }, { "epoch": 2.9575156868230685, "grad_norm": 0.1217394471168518, "learning_rate": 1.530240402663502e-06, "loss": 0.0234, "step": 148000 }, { "epoch": 2.9577155189640703, "grad_norm": 0.08306468278169632, "learning_rate": 1.5277043381237478e-06, "loss": 0.0225, "step": 148010 }, { "epoch": 2.957915351105072, "grad_norm": 0.24614828824996948, "learning_rate": 1.5251703442314192e-06, "loss": 0.0222, "step": 148020 }, { "epoch": 2.9581151832460733, "grad_norm": 0.11785628646612167, "learning_rate": 1.522638421094763e-06, "loss": 0.028, "step": 148030 }, { "epoch": 2.958315015387075, "grad_norm": 0.11936185508966446, "learning_rate": 1.5201085688219375e-06, "loss": 0.0237, "step": 148040 }, { "epoch": 2.9585148475280763, "grad_norm": 0.10555926710367203, "learning_rate": 1.5175807875210112e-06, "loss": 0.0186, "step": 148050 }, { "epoch": 2.958714679669078, "grad_norm": 0.13717374205589294, "learning_rate": 1.5150550772999705e-06, "loss": 0.0228, "step": 148060 }, { "epoch": 2.9589145118100797, "grad_norm": 0.12298893928527832, "learning_rate": 1.5125314382667066e-06, "loss": 0.0186, "step": 148070 }, { "epoch": 2.959114343951081, "grad_norm": 0.10802972316741943, "learning_rate": 1.5100098705290222e-06, "loss": 0.0249, "step": 148080 }, { "epoch": 2.9593141760920827, "grad_norm": 0.13057665526866913, "learning_rate": 1.5074903741946477e-06, "loss": 0.0247, "step": 148090 }, { "epoch": 2.959514008233084, "grad_norm": 0.09244078397750854, "learning_rate": 1.5049729493711862e-06, "loss": 0.0245, "step": 148100 }, { "epoch": 2.9597138403740857, "grad_norm": 0.13551756739616394, "learning_rate": 1.5024575961662013e-06, "loss": 0.0239, "step": 148110 }, { "epoch": 2.9599136725150874, "grad_norm": 0.07780196517705917, "learning_rate": 1.499944314687135e-06, "loss": 0.0231, "step": 148120 }, { "epoch": 2.960113504656089, "grad_norm": 0.15449054539203644, "learning_rate": 1.4974331050413515e-06, "loss": 0.0306, "step": 148130 }, { "epoch": 2.9603133367970904, "grad_norm": 0.10054904967546463, "learning_rate": 1.4949239673361148e-06, "loss": 0.0293, "step": 148140 }, { "epoch": 2.960513168938092, "grad_norm": 0.1812421828508377, "learning_rate": 1.4924169016786283e-06, "loss": 0.0309, "step": 148150 }, { "epoch": 2.9607130010790934, "grad_norm": 0.15732644498348236, "learning_rate": 1.4899119081759782e-06, "loss": 0.0296, "step": 148160 }, { "epoch": 2.960912833220095, "grad_norm": 0.09708339720964432, "learning_rate": 1.4874089869351736e-06, "loss": 0.0249, "step": 148170 }, { "epoch": 2.961112665361097, "grad_norm": 0.1250772476196289, "learning_rate": 1.4849081380631347e-06, "loss": 0.0232, "step": 148180 }, { "epoch": 2.961312497502098, "grad_norm": 0.10903669148683548, "learning_rate": 1.482409361666698e-06, "loss": 0.0164, "step": 148190 }, { "epoch": 2.9615123296431, "grad_norm": 0.08081098645925522, "learning_rate": 1.4799126578526057e-06, "loss": 0.029, "step": 148200 }, { "epoch": 2.961712161784101, "grad_norm": 0.14987383782863617, "learning_rate": 1.4774180267275062e-06, "loss": 0.0265, "step": 148210 }, { "epoch": 2.961911993925103, "grad_norm": 0.11394548416137695, "learning_rate": 1.4749254683979807e-06, "loss": 0.0259, "step": 148220 }, { "epoch": 2.9621118260661046, "grad_norm": 0.14776553213596344, "learning_rate": 1.4724349829704775e-06, "loss": 0.0239, "step": 148230 }, { "epoch": 2.9623116582071063, "grad_norm": 0.14404179155826569, "learning_rate": 1.469946570551417e-06, "loss": 0.0297, "step": 148240 }, { "epoch": 2.9625114903481076, "grad_norm": 0.0987146645784378, "learning_rate": 1.4674602312470865e-06, "loss": 0.0314, "step": 148250 }, { "epoch": 2.9627113224891093, "grad_norm": 0.16243746876716614, "learning_rate": 1.4649759651636952e-06, "loss": 0.0287, "step": 148260 }, { "epoch": 2.9629111546301106, "grad_norm": 0.09613117575645447, "learning_rate": 1.462493772407364e-06, "loss": 0.0214, "step": 148270 }, { "epoch": 2.9631109867711123, "grad_norm": 0.091868095099926, "learning_rate": 1.4600136530841356e-06, "loss": 0.0248, "step": 148280 }, { "epoch": 2.963310818912114, "grad_norm": 0.12172922492027283, "learning_rate": 1.4575356072999535e-06, "loss": 0.0239, "step": 148290 }, { "epoch": 2.9635106510531153, "grad_norm": 0.14526653289794922, "learning_rate": 1.4550596351606771e-06, "loss": 0.0274, "step": 148300 }, { "epoch": 2.963710483194117, "grad_norm": 0.09833091497421265, "learning_rate": 1.4525857367720718e-06, "loss": 0.0194, "step": 148310 }, { "epoch": 2.9639103153351183, "grad_norm": 0.07702402025461197, "learning_rate": 1.45011391223982e-06, "loss": 0.0227, "step": 148320 }, { "epoch": 2.96411014747612, "grad_norm": 0.10832176357507706, "learning_rate": 1.4476441616695092e-06, "loss": 0.0196, "step": 148330 }, { "epoch": 2.9643099796171217, "grad_norm": 0.08818433433771133, "learning_rate": 1.4451764851666494e-06, "loss": 0.0246, "step": 148340 }, { "epoch": 2.9645098117581234, "grad_norm": 0.1026172935962677, "learning_rate": 1.4427108828366564e-06, "loss": 0.0224, "step": 148350 }, { "epoch": 2.9647096438991247, "grad_norm": 0.12693072855472565, "learning_rate": 1.4402473547848462e-06, "loss": 0.0266, "step": 148360 }, { "epoch": 2.9649094760401264, "grad_norm": 0.12466377019882202, "learning_rate": 1.4377859011164618e-06, "loss": 0.0264, "step": 148370 }, { "epoch": 2.9651093081811277, "grad_norm": 0.11412665247917175, "learning_rate": 1.4353265219366586e-06, "loss": 0.0279, "step": 148380 }, { "epoch": 2.9653091403221294, "grad_norm": 0.08592195808887482, "learning_rate": 1.43286921735048e-06, "loss": 0.0202, "step": 148390 }, { "epoch": 2.965508972463131, "grad_norm": 0.13280344009399414, "learning_rate": 1.430413987462914e-06, "loss": 0.0247, "step": 148400 }, { "epoch": 2.9657088046041324, "grad_norm": 0.10854722559452057, "learning_rate": 1.4279608323788329e-06, "loss": 0.0201, "step": 148410 }, { "epoch": 2.965908636745134, "grad_norm": 0.16686579585075378, "learning_rate": 1.4255097522030414e-06, "loss": 0.0199, "step": 148420 }, { "epoch": 2.9661084688861354, "grad_norm": 0.12086500227451324, "learning_rate": 1.4230607470402334e-06, "loss": 0.024, "step": 148430 }, { "epoch": 2.966308301027137, "grad_norm": 0.12893064320087433, "learning_rate": 1.4206138169950366e-06, "loss": 0.0205, "step": 148440 }, { "epoch": 2.966508133168139, "grad_norm": 0.18996264040470123, "learning_rate": 1.418168962171973e-06, "loss": 0.0237, "step": 148450 }, { "epoch": 2.96670796530914, "grad_norm": 0.10170447826385498, "learning_rate": 1.4157261826754808e-06, "loss": 0.0279, "step": 148460 }, { "epoch": 2.966907797450142, "grad_norm": 0.11456992477178574, "learning_rate": 1.413285478609916e-06, "loss": 0.0221, "step": 148470 }, { "epoch": 2.9671076295911436, "grad_norm": 0.13488468527793884, "learning_rate": 1.4108468500795392e-06, "loss": 0.0276, "step": 148480 }, { "epoch": 2.967307461732145, "grad_norm": 0.13309460878372192, "learning_rate": 1.4084102971885283e-06, "loss": 0.0237, "step": 148490 }, { "epoch": 2.9675072938731466, "grad_norm": 0.10343852639198303, "learning_rate": 1.4059758200409502e-06, "loss": 0.0234, "step": 148500 }, { "epoch": 2.9677071260141483, "grad_norm": 0.094539575278759, "learning_rate": 1.403543418740827e-06, "loss": 0.0286, "step": 148510 }, { "epoch": 2.9679069581551496, "grad_norm": 0.10923107713460922, "learning_rate": 1.4011130933920536e-06, "loss": 0.0272, "step": 148520 }, { "epoch": 2.9681067902961513, "grad_norm": 0.12361781299114227, "learning_rate": 1.3986848440984467e-06, "loss": 0.0246, "step": 148530 }, { "epoch": 2.9683066224371526, "grad_norm": 0.1378546804189682, "learning_rate": 1.3962586709637404e-06, "loss": 0.0213, "step": 148540 }, { "epoch": 2.9685064545781543, "grad_norm": 0.15047018229961395, "learning_rate": 1.3938345740915737e-06, "loss": 0.0166, "step": 148550 }, { "epoch": 2.968706286719156, "grad_norm": 0.08153253048658371, "learning_rate": 1.3914125535855027e-06, "loss": 0.024, "step": 148560 }, { "epoch": 2.9689061188601573, "grad_norm": 0.09565906226634979, "learning_rate": 1.3889926095489946e-06, "loss": 0.0247, "step": 148570 }, { "epoch": 2.969105951001159, "grad_norm": 0.1159970834851265, "learning_rate": 1.386574742085417e-06, "loss": 0.0204, "step": 148580 }, { "epoch": 2.9693057831421603, "grad_norm": 0.15885725617408752, "learning_rate": 1.3841589512980591e-06, "loss": 0.0212, "step": 148590 }, { "epoch": 2.969505615283162, "grad_norm": 0.11882650852203369, "learning_rate": 1.3817452372901273e-06, "loss": 0.0208, "step": 148600 }, { "epoch": 2.9697054474241638, "grad_norm": 0.12237906455993652, "learning_rate": 1.3793336001647172e-06, "loss": 0.0283, "step": 148610 }, { "epoch": 2.9699052795651655, "grad_norm": 0.12564347684383392, "learning_rate": 1.3769240400248628e-06, "loss": 0.0278, "step": 148620 }, { "epoch": 2.9701051117061668, "grad_norm": 0.1146959662437439, "learning_rate": 1.3745165569734874e-06, "loss": 0.027, "step": 148630 }, { "epoch": 2.9703049438471685, "grad_norm": 0.09127512574195862, "learning_rate": 1.3721111511134311e-06, "loss": 0.0239, "step": 148640 }, { "epoch": 2.9705047759881698, "grad_norm": 0.12121317535638809, "learning_rate": 1.3697078225474668e-06, "loss": 0.0218, "step": 148650 }, { "epoch": 2.9707046081291715, "grad_norm": 0.13037127256393433, "learning_rate": 1.3673065713782407e-06, "loss": 0.0197, "step": 148660 }, { "epoch": 2.970904440270173, "grad_norm": 0.11652746051549911, "learning_rate": 1.3649073977083316e-06, "loss": 0.0274, "step": 148670 }, { "epoch": 2.9711042724111745, "grad_norm": 0.08168434351682663, "learning_rate": 1.3625103016402407e-06, "loss": 0.0182, "step": 148680 }, { "epoch": 2.971304104552176, "grad_norm": 0.10797243565320969, "learning_rate": 1.3601152832763531e-06, "loss": 0.0185, "step": 148690 }, { "epoch": 2.9715039366931775, "grad_norm": 0.0900648757815361, "learning_rate": 1.3577223427189921e-06, "loss": 0.0273, "step": 148700 }, { "epoch": 2.971703768834179, "grad_norm": 0.15007933974266052, "learning_rate": 1.3553314800703709e-06, "loss": 0.0251, "step": 148710 }, { "epoch": 2.971903600975181, "grad_norm": 0.20711860060691833, "learning_rate": 1.3529426954326297e-06, "loss": 0.029, "step": 148720 }, { "epoch": 2.9721034331161826, "grad_norm": 0.14193059504032135, "learning_rate": 1.3505559889078034e-06, "loss": 0.0248, "step": 148730 }, { "epoch": 2.972303265257184, "grad_norm": 0.09959319233894348, "learning_rate": 1.3481713605978553e-06, "loss": 0.023, "step": 148740 }, { "epoch": 2.9725030973981856, "grad_norm": 0.1536005288362503, "learning_rate": 1.3457888106046534e-06, "loss": 0.0296, "step": 148750 }, { "epoch": 2.972702929539187, "grad_norm": 0.08468509465456009, "learning_rate": 1.3434083390299723e-06, "loss": 0.0268, "step": 148760 }, { "epoch": 2.9729027616801886, "grad_norm": 0.10885636508464813, "learning_rate": 1.3410299459754917e-06, "loss": 0.0223, "step": 148770 }, { "epoch": 2.9731025938211904, "grad_norm": 0.10467337816953659, "learning_rate": 1.3386536315428355e-06, "loss": 0.0215, "step": 148780 }, { "epoch": 2.9733024259621916, "grad_norm": 0.09620849043130875, "learning_rate": 1.3362793958335007e-06, "loss": 0.0169, "step": 148790 }, { "epoch": 2.9735022581031934, "grad_norm": 0.15510867536067963, "learning_rate": 1.333907238948906e-06, "loss": 0.0275, "step": 148800 }, { "epoch": 2.9737020902441946, "grad_norm": 0.08759056776762009, "learning_rate": 1.3315371609903981e-06, "loss": 0.0315, "step": 148810 }, { "epoch": 2.9739019223851964, "grad_norm": 0.0935019850730896, "learning_rate": 1.3291691620592072e-06, "loss": 0.0245, "step": 148820 }, { "epoch": 2.974101754526198, "grad_norm": 0.10141264647245407, "learning_rate": 1.3268032422565079e-06, "loss": 0.0181, "step": 148830 }, { "epoch": 2.9743015866672, "grad_norm": 0.11041145026683807, "learning_rate": 1.3244394016833528e-06, "loss": 0.0168, "step": 148840 }, { "epoch": 2.974501418808201, "grad_norm": 0.13558870553970337, "learning_rate": 1.3220776404407276e-06, "loss": 0.022, "step": 148850 }, { "epoch": 2.974701250949203, "grad_norm": 0.11937298625707626, "learning_rate": 1.3197179586295182e-06, "loss": 0.0243, "step": 148860 }, { "epoch": 2.974901083090204, "grad_norm": 0.11351180821657181, "learning_rate": 1.3173603563505332e-06, "loss": 0.0262, "step": 148870 }, { "epoch": 2.975100915231206, "grad_norm": 0.12489044666290283, "learning_rate": 1.3150048337044807e-06, "loss": 0.0251, "step": 148880 }, { "epoch": 2.9753007473722075, "grad_norm": 0.09001918137073517, "learning_rate": 1.3126513907919913e-06, "loss": 0.0247, "step": 148890 }, { "epoch": 2.975500579513209, "grad_norm": 0.10401143878698349, "learning_rate": 1.3103000277135846e-06, "loss": 0.0225, "step": 148900 }, { "epoch": 2.9757004116542105, "grad_norm": 0.16479763388633728, "learning_rate": 1.3079507445697136e-06, "loss": 0.0246, "step": 148910 }, { "epoch": 2.975900243795212, "grad_norm": 0.12292244285345078, "learning_rate": 1.3056035414607426e-06, "loss": 0.0244, "step": 148920 }, { "epoch": 2.9761000759362135, "grad_norm": 0.14998720586299896, "learning_rate": 1.30325841848693e-06, "loss": 0.0234, "step": 148930 }, { "epoch": 2.9762999080772152, "grad_norm": 0.14306631684303284, "learning_rate": 1.300915375748457e-06, "loss": 0.0254, "step": 148940 }, { "epoch": 2.976499740218217, "grad_norm": 0.11218668520450592, "learning_rate": 1.2985744133454215e-06, "loss": 0.0288, "step": 148950 }, { "epoch": 2.9766995723592182, "grad_norm": 0.14962343871593475, "learning_rate": 1.2962355313778152e-06, "loss": 0.0237, "step": 148960 }, { "epoch": 2.97689940450022, "grad_norm": 0.11512979865074158, "learning_rate": 1.2938987299455585e-06, "loss": 0.0228, "step": 148970 }, { "epoch": 2.9770992366412212, "grad_norm": 0.11400553584098816, "learning_rate": 1.291564009148466e-06, "loss": 0.0196, "step": 148980 }, { "epoch": 2.977299068782223, "grad_norm": 0.12069939076900482, "learning_rate": 1.2892313690862911e-06, "loss": 0.0223, "step": 148990 }, { "epoch": 2.9774989009232247, "grad_norm": 0.18790996074676514, "learning_rate": 1.2869008098586544e-06, "loss": 0.026, "step": 149000 }, { "epoch": 2.977698733064226, "grad_norm": 0.12110647559165955, "learning_rate": 1.2845723315651315e-06, "loss": 0.0248, "step": 149010 }, { "epoch": 2.9778985652052277, "grad_norm": 0.09675120562314987, "learning_rate": 1.2822459343051874e-06, "loss": 0.0242, "step": 149020 }, { "epoch": 2.978098397346229, "grad_norm": 0.114219531416893, "learning_rate": 1.279921618178198e-06, "loss": 0.0308, "step": 149030 }, { "epoch": 2.9782982294872307, "grad_norm": 0.10235504806041718, "learning_rate": 1.277599383283451e-06, "loss": 0.022, "step": 149040 }, { "epoch": 2.9784980616282324, "grad_norm": 0.16882386803627014, "learning_rate": 1.275279229720161e-06, "loss": 0.0208, "step": 149050 }, { "epoch": 2.978697893769234, "grad_norm": 0.12109378725290298, "learning_rate": 1.2729611575874268e-06, "loss": 0.0252, "step": 149060 }, { "epoch": 2.9788977259102354, "grad_norm": 0.14078792929649353, "learning_rate": 1.2706451669842801e-06, "loss": 0.026, "step": 149070 }, { "epoch": 2.979097558051237, "grad_norm": 0.12360271066427231, "learning_rate": 1.2683312580096529e-06, "loss": 0.0188, "step": 149080 }, { "epoch": 2.9792973901922384, "grad_norm": 0.13112083077430725, "learning_rate": 1.2660194307623885e-06, "loss": 0.0244, "step": 149090 }, { "epoch": 2.97949722233324, "grad_norm": 0.1477459818124771, "learning_rate": 1.2637096853412467e-06, "loss": 0.0273, "step": 149100 }, { "epoch": 2.979697054474242, "grad_norm": 0.1291537582874298, "learning_rate": 1.2614020218448986e-06, "loss": 0.0223, "step": 149110 }, { "epoch": 2.979896886615243, "grad_norm": 0.1052817851305008, "learning_rate": 1.2590964403719207e-06, "loss": 0.0225, "step": 149120 }, { "epoch": 2.980096718756245, "grad_norm": 0.09851056337356567, "learning_rate": 1.2567929410207956e-06, "loss": 0.0232, "step": 149130 }, { "epoch": 2.980296550897246, "grad_norm": 0.12127446383237839, "learning_rate": 1.254491523889939e-06, "loss": 0.0274, "step": 149140 }, { "epoch": 2.980496383038248, "grad_norm": 0.11481639742851257, "learning_rate": 1.252192189077661e-06, "loss": 0.0223, "step": 149150 }, { "epoch": 2.9806962151792495, "grad_norm": 0.11428288370370865, "learning_rate": 1.2498949366821721e-06, "loss": 0.0366, "step": 149160 }, { "epoch": 2.9808960473202513, "grad_norm": 0.15847574174404144, "learning_rate": 1.247599766801616e-06, "loss": 0.0244, "step": 149170 }, { "epoch": 2.9810958794612525, "grad_norm": 0.09070136398077011, "learning_rate": 1.2453066795340362e-06, "loss": 0.0213, "step": 149180 }, { "epoch": 2.9812957116022543, "grad_norm": 0.12527665495872498, "learning_rate": 1.2430156749773935e-06, "loss": 0.0235, "step": 149190 }, { "epoch": 2.9814955437432555, "grad_norm": 0.0918341875076294, "learning_rate": 1.2407267532295485e-06, "loss": 0.0241, "step": 149200 }, { "epoch": 2.9816953758842573, "grad_norm": 0.08758345246315002, "learning_rate": 1.2384399143882842e-06, "loss": 0.0244, "step": 149210 }, { "epoch": 2.981895208025259, "grad_norm": 0.12095402926206589, "learning_rate": 1.2361551585512887e-06, "loss": 0.0254, "step": 149220 }, { "epoch": 2.9820950401662603, "grad_norm": 0.14769187569618225, "learning_rate": 1.2338724858161622e-06, "loss": 0.0229, "step": 149230 }, { "epoch": 2.982294872307262, "grad_norm": 0.1573973298072815, "learning_rate": 1.2315918962804152e-06, "loss": 0.0222, "step": 149240 }, { "epoch": 2.9824947044482633, "grad_norm": 0.12221860885620117, "learning_rate": 1.2293133900414755e-06, "loss": 0.0226, "step": 149250 }, { "epoch": 2.982694536589265, "grad_norm": 0.14087699353694916, "learning_rate": 1.227036967196671e-06, "loss": 0.0256, "step": 149260 }, { "epoch": 2.9828943687302667, "grad_norm": 0.09341853857040405, "learning_rate": 1.2247626278432511e-06, "loss": 0.0258, "step": 149270 }, { "epoch": 2.983094200871268, "grad_norm": 0.12423252314329147, "learning_rate": 1.2224903720783664e-06, "loss": 0.0201, "step": 149280 }, { "epoch": 2.9832940330122697, "grad_norm": 0.17776019871234894, "learning_rate": 1.2202201999990892e-06, "loss": 0.0211, "step": 149290 }, { "epoch": 2.9834938651532714, "grad_norm": 0.11287106573581696, "learning_rate": 1.2179521117023918e-06, "loss": 0.0225, "step": 149300 }, { "epoch": 2.9836936972942727, "grad_norm": 0.08341699838638306, "learning_rate": 1.2156861072851578e-06, "loss": 0.0264, "step": 149310 }, { "epoch": 2.9838935294352744, "grad_norm": 0.11990430951118469, "learning_rate": 1.2134221868442042e-06, "loss": 0.0271, "step": 149320 }, { "epoch": 2.984093361576276, "grad_norm": 0.13057467341423035, "learning_rate": 1.211160350476226e-06, "loss": 0.0203, "step": 149330 }, { "epoch": 2.9842931937172774, "grad_norm": 0.21642570197582245, "learning_rate": 1.2089005982778512e-06, "loss": 0.021, "step": 149340 }, { "epoch": 2.984493025858279, "grad_norm": 0.1261759251356125, "learning_rate": 1.2066429303456084e-06, "loss": 0.0302, "step": 149350 }, { "epoch": 2.9846928579992804, "grad_norm": 0.11076721549034119, "learning_rate": 1.2043873467759426e-06, "loss": 0.0255, "step": 149360 }, { "epoch": 2.984892690140282, "grad_norm": 0.09239191561937332, "learning_rate": 1.2021338476652156e-06, "loss": 0.0235, "step": 149370 }, { "epoch": 2.985092522281284, "grad_norm": 0.10178348422050476, "learning_rate": 1.1998824331096781e-06, "loss": 0.0206, "step": 149380 }, { "epoch": 2.985292354422285, "grad_norm": 0.12852418422698975, "learning_rate": 1.1976331032055254e-06, "loss": 0.0257, "step": 149390 }, { "epoch": 2.985492186563287, "grad_norm": 0.143077090382576, "learning_rate": 1.1953858580488198e-06, "loss": 0.0258, "step": 149400 }, { "epoch": 2.985692018704288, "grad_norm": 0.12914958596229553, "learning_rate": 1.1931406977355786e-06, "loss": 0.026, "step": 149410 }, { "epoch": 2.98589185084529, "grad_norm": 0.11018259823322296, "learning_rate": 1.1908976223617086e-06, "loss": 0.0222, "step": 149420 }, { "epoch": 2.9860916829862916, "grad_norm": 0.12731914222240448, "learning_rate": 1.1886566320230274e-06, "loss": 0.0194, "step": 149430 }, { "epoch": 2.9862915151272933, "grad_norm": 0.130193829536438, "learning_rate": 1.1864177268152587e-06, "loss": 0.024, "step": 149440 }, { "epoch": 2.9864913472682946, "grad_norm": 0.15274769067764282, "learning_rate": 1.1841809068340592e-06, "loss": 0.0222, "step": 149450 }, { "epoch": 2.9866911794092963, "grad_norm": 0.2039841264486313, "learning_rate": 1.1819461721749691e-06, "loss": 0.0255, "step": 149460 }, { "epoch": 2.9868910115502976, "grad_norm": 0.1257386952638626, "learning_rate": 1.1797135229334566e-06, "loss": 0.0289, "step": 149470 }, { "epoch": 2.9870908436912993, "grad_norm": 0.12282867729663849, "learning_rate": 1.1774829592049007e-06, "loss": 0.0215, "step": 149480 }, { "epoch": 2.987290675832301, "grad_norm": 0.13222377002239227, "learning_rate": 1.175254481084581e-06, "loss": 0.0233, "step": 149490 }, { "epoch": 2.9874905079733023, "grad_norm": 0.09288687258958817, "learning_rate": 1.1730280886676937e-06, "loss": 0.0226, "step": 149500 }, { "epoch": 2.987690340114304, "grad_norm": 0.12404672801494598, "learning_rate": 1.1708037820493511e-06, "loss": 0.0289, "step": 149510 }, { "epoch": 2.9878901722553053, "grad_norm": 0.07076758146286011, "learning_rate": 1.168581561324572e-06, "loss": 0.0223, "step": 149520 }, { "epoch": 2.988090004396307, "grad_norm": 0.12096837908029556, "learning_rate": 1.166361426588275e-06, "loss": 0.0192, "step": 149530 }, { "epoch": 2.9882898365373087, "grad_norm": 0.1186177060008049, "learning_rate": 1.1641433779353117e-06, "loss": 0.0253, "step": 149540 }, { "epoch": 2.9884896686783105, "grad_norm": 0.1175176128745079, "learning_rate": 1.1619274154604288e-06, "loss": 0.0227, "step": 149550 }, { "epoch": 2.9886895008193117, "grad_norm": 0.12184864282608032, "learning_rate": 1.1597135392582891e-06, "loss": 0.0165, "step": 149560 }, { "epoch": 2.9888893329603134, "grad_norm": 0.11163385957479477, "learning_rate": 1.1575017494234619e-06, "loss": 0.0259, "step": 149570 }, { "epoch": 2.9890891651013147, "grad_norm": 0.14696837961673737, "learning_rate": 1.1552920460504325e-06, "loss": 0.0247, "step": 149580 }, { "epoch": 2.9892889972423164, "grad_norm": 0.08804082870483398, "learning_rate": 1.153084429233603e-06, "loss": 0.0282, "step": 149590 }, { "epoch": 2.989488829383318, "grad_norm": 0.11348118633031845, "learning_rate": 1.1508788990672647e-06, "loss": 0.0195, "step": 149600 }, { "epoch": 2.9896886615243194, "grad_norm": 0.1399317979812622, "learning_rate": 1.1486754556456425e-06, "loss": 0.0254, "step": 149610 }, { "epoch": 2.989888493665321, "grad_norm": 0.16775432229042053, "learning_rate": 1.1464740990628665e-06, "loss": 0.0228, "step": 149620 }, { "epoch": 2.9900883258063224, "grad_norm": 0.11501513421535492, "learning_rate": 1.1442748294129667e-06, "loss": 0.0207, "step": 149630 }, { "epoch": 2.990288157947324, "grad_norm": 0.11824308335781097, "learning_rate": 1.1420776467898964e-06, "loss": 0.0291, "step": 149640 }, { "epoch": 2.990487990088326, "grad_norm": 0.13437345623970032, "learning_rate": 1.1398825512875133e-06, "loss": 0.0172, "step": 149650 }, { "epoch": 2.9906878222293276, "grad_norm": 0.1231650784611702, "learning_rate": 1.1376895429995926e-06, "loss": 0.0275, "step": 149660 }, { "epoch": 2.990887654370329, "grad_norm": 0.12496891617774963, "learning_rate": 1.135498622019804e-06, "loss": 0.0298, "step": 149670 }, { "epoch": 2.9910874865113306, "grad_norm": 0.11552048474550247, "learning_rate": 1.1333097884417499e-06, "loss": 0.0303, "step": 149680 }, { "epoch": 2.991287318652332, "grad_norm": 0.06803574413061142, "learning_rate": 1.1311230423589391e-06, "loss": 0.0185, "step": 149690 }, { "epoch": 2.9914871507933336, "grad_norm": 0.14409200847148895, "learning_rate": 1.1289383838647693e-06, "loss": 0.0225, "step": 149700 }, { "epoch": 2.9916869829343353, "grad_norm": 0.2419711947441101, "learning_rate": 1.1267558130525657e-06, "loss": 0.025, "step": 149710 }, { "epoch": 2.9918868150753366, "grad_norm": 0.11984066665172577, "learning_rate": 1.124575330015587e-06, "loss": 0.0201, "step": 149720 }, { "epoch": 2.9920866472163383, "grad_norm": 0.10140577703714371, "learning_rate": 1.1223969348469533e-06, "loss": 0.0219, "step": 149730 }, { "epoch": 2.9922864793573396, "grad_norm": 0.12468265742063522, "learning_rate": 1.1202206276397287e-06, "loss": 0.0233, "step": 149740 }, { "epoch": 2.9924863114983413, "grad_norm": 0.11399564146995544, "learning_rate": 1.118046408486889e-06, "loss": 0.026, "step": 149750 }, { "epoch": 2.992686143639343, "grad_norm": 0.11386143416166306, "learning_rate": 1.1158742774813103e-06, "loss": 0.0245, "step": 149760 }, { "epoch": 2.9928859757803448, "grad_norm": 0.12879379093647003, "learning_rate": 1.1137042347157733e-06, "loss": 0.0281, "step": 149770 }, { "epoch": 2.993085807921346, "grad_norm": 0.12777207791805267, "learning_rate": 1.1115362802829875e-06, "loss": 0.0295, "step": 149780 }, { "epoch": 2.9932856400623478, "grad_norm": 0.1152055487036705, "learning_rate": 1.1093704142755622e-06, "loss": 0.0229, "step": 149790 }, { "epoch": 2.993485472203349, "grad_norm": 0.3199091851711273, "learning_rate": 1.1072066367860124e-06, "loss": 0.0259, "step": 149800 }, { "epoch": 2.9936853043443508, "grad_norm": 0.08961082249879837, "learning_rate": 1.1050449479067804e-06, "loss": 0.0247, "step": 149810 }, { "epoch": 2.9938851364853525, "grad_norm": 0.14981043338775635, "learning_rate": 1.1028853477302093e-06, "loss": 0.0278, "step": 149820 }, { "epoch": 2.9940849686263538, "grad_norm": 0.1887025088071823, "learning_rate": 1.1007278363485419e-06, "loss": 0.0286, "step": 149830 }, { "epoch": 2.9942848007673555, "grad_norm": 0.11623018234968185, "learning_rate": 1.0985724138539544e-06, "loss": 0.0204, "step": 149840 }, { "epoch": 2.9944846329083568, "grad_norm": 0.12287195026874542, "learning_rate": 1.0964190803385177e-06, "loss": 0.0218, "step": 149850 }, { "epoch": 2.9946844650493585, "grad_norm": 0.12167021632194519, "learning_rate": 1.0942678358942194e-06, "loss": 0.0279, "step": 149860 }, { "epoch": 2.99488429719036, "grad_norm": 0.0991973951458931, "learning_rate": 1.092118680612958e-06, "loss": 0.0218, "step": 149870 }, { "epoch": 2.995084129331362, "grad_norm": 0.11580205708742142, "learning_rate": 1.089971614586538e-06, "loss": 0.023, "step": 149880 }, { "epoch": 2.995283961472363, "grad_norm": 0.14905822277069092, "learning_rate": 1.0878266379066748e-06, "loss": 0.0269, "step": 149890 }, { "epoch": 2.995483793613365, "grad_norm": 0.13576005399227142, "learning_rate": 1.0856837506650119e-06, "loss": 0.0208, "step": 149900 }, { "epoch": 2.995683625754366, "grad_norm": 0.14637398719787598, "learning_rate": 1.083542952953076e-06, "loss": 0.0168, "step": 149910 }, { "epoch": 2.995883457895368, "grad_norm": 0.10066229850053787, "learning_rate": 1.0814042448623275e-06, "loss": 0.0213, "step": 149920 }, { "epoch": 2.9960832900363696, "grad_norm": 0.18132847547531128, "learning_rate": 1.0792676264841207e-06, "loss": 0.027, "step": 149930 }, { "epoch": 2.996283122177371, "grad_norm": 0.11540599167346954, "learning_rate": 1.0771330979097216e-06, "loss": 0.0227, "step": 149940 }, { "epoch": 2.9964829543183726, "grad_norm": 0.07798024266958237, "learning_rate": 1.0750006592303352e-06, "loss": 0.0245, "step": 149950 }, { "epoch": 2.996682786459374, "grad_norm": 0.08652178943157196, "learning_rate": 1.0728703105370386e-06, "loss": 0.0255, "step": 149960 }, { "epoch": 2.9968826186003756, "grad_norm": 0.1364152878522873, "learning_rate": 1.0707420519208366e-06, "loss": 0.0339, "step": 149970 }, { "epoch": 2.9970824507413774, "grad_norm": 0.10906539857387543, "learning_rate": 1.0686158834726456e-06, "loss": 0.0256, "step": 149980 }, { "epoch": 2.997282282882379, "grad_norm": 0.10527438670396805, "learning_rate": 1.066491805283304e-06, "loss": 0.0248, "step": 149990 }, { "epoch": 2.9974821150233804, "grad_norm": 0.1593768298625946, "learning_rate": 1.064369817443539e-06, "loss": 0.0233, "step": 150000 }, { "epoch": 2.997681947164382, "grad_norm": 0.09078546613454819, "learning_rate": 1.0622499200439894e-06, "loss": 0.0289, "step": 150010 }, { "epoch": 2.9978817793053834, "grad_norm": 0.11817052215337753, "learning_rate": 1.0601321131752273e-06, "loss": 0.0315, "step": 150020 }, { "epoch": 2.998081611446385, "grad_norm": 0.11595719307661057, "learning_rate": 1.0580163969277135e-06, "loss": 0.0275, "step": 150030 }, { "epoch": 2.998281443587387, "grad_norm": 0.14339421689510345, "learning_rate": 1.0559027713918312e-06, "loss": 0.03, "step": 150040 }, { "epoch": 2.998481275728388, "grad_norm": 0.08007130771875381, "learning_rate": 1.0537912366578695e-06, "loss": 0.0246, "step": 150050 }, { "epoch": 2.99868110786939, "grad_norm": 0.1264786720275879, "learning_rate": 1.0516817928160282e-06, "loss": 0.0214, "step": 150060 }, { "epoch": 2.998880940010391, "grad_norm": 0.12464304268360138, "learning_rate": 1.049574439956419e-06, "loss": 0.0299, "step": 150070 }, { "epoch": 2.999080772151393, "grad_norm": 0.11082279682159424, "learning_rate": 1.0474691781690637e-06, "loss": 0.0312, "step": 150080 }, { "epoch": 2.9992806042923945, "grad_norm": 0.11541217565536499, "learning_rate": 1.0453660075439021e-06, "loss": 0.025, "step": 150090 }, { "epoch": 2.999480436433396, "grad_norm": 0.13160757720470428, "learning_rate": 1.0432649281707674e-06, "loss": 0.032, "step": 150100 }, { "epoch": 2.9996802685743975, "grad_norm": 0.14456701278686523, "learning_rate": 1.0411659401394214e-06, "loss": 0.022, "step": 150110 }, { "epoch": 2.999880100715399, "grad_norm": 0.13924941420555115, "learning_rate": 1.0390690435395257e-06, "loss": 0.0289, "step": 150120 }, { "epoch": 3.0000799328564005, "grad_norm": 0.18022596836090088, "learning_rate": 1.0369742384606529e-06, "loss": 0.0255, "step": 150130 }, { "epoch": 3.0002797649974022, "grad_norm": 0.10153494030237198, "learning_rate": 1.0348815249922927e-06, "loss": 0.0275, "step": 150140 }, { "epoch": 3.000479597138404, "grad_norm": 0.11935127526521683, "learning_rate": 1.0327909032238458e-06, "loss": 0.0254, "step": 150150 }, { "epoch": 3.0006794292794052, "grad_norm": 0.12306995689868927, "learning_rate": 1.0307023732446131e-06, "loss": 0.0225, "step": 150160 }, { "epoch": 3.000879261420407, "grad_norm": 0.16882172226905823, "learning_rate": 1.0286159351438174e-06, "loss": 0.0252, "step": 150170 }, { "epoch": 3.0010790935614082, "grad_norm": 0.09823360294103622, "learning_rate": 1.026531589010582e-06, "loss": 0.0236, "step": 150180 }, { "epoch": 3.00127892570241, "grad_norm": 0.10522152483463287, "learning_rate": 1.024449334933958e-06, "loss": 0.0205, "step": 150190 }, { "epoch": 3.0014787578434117, "grad_norm": 0.08980926871299744, "learning_rate": 1.022369173002885e-06, "loss": 0.0185, "step": 150200 }, { "epoch": 3.001678589984413, "grad_norm": 0.0957564264535904, "learning_rate": 1.0202911033062145e-06, "loss": 0.0229, "step": 150210 }, { "epoch": 3.0018784221254147, "grad_norm": 0.1372997760772705, "learning_rate": 1.018215125932742e-06, "loss": 0.0207, "step": 150220 }, { "epoch": 3.0020782542664164, "grad_norm": 0.11977649480104446, "learning_rate": 1.0161412409711358e-06, "loss": 0.022, "step": 150230 }, { "epoch": 3.0022780864074177, "grad_norm": 0.10099858045578003, "learning_rate": 1.0140694485099855e-06, "loss": 0.0364, "step": 150240 }, { "epoch": 3.0024779185484194, "grad_norm": 0.09742086380720139, "learning_rate": 1.0119997486377986e-06, "loss": 0.0186, "step": 150250 }, { "epoch": 3.0026777506894207, "grad_norm": 0.11783968657255173, "learning_rate": 1.009932141442993e-06, "loss": 0.0312, "step": 150260 }, { "epoch": 3.0028775828304224, "grad_norm": 0.13572822511196136, "learning_rate": 1.007866627013887e-06, "loss": 0.0265, "step": 150270 }, { "epoch": 3.003077414971424, "grad_norm": 0.08024084568023682, "learning_rate": 1.0058032054387157e-06, "loss": 0.0221, "step": 150280 }, { "epoch": 3.0032772471124254, "grad_norm": 0.1144244447350502, "learning_rate": 1.0037418768056305e-06, "loss": 0.026, "step": 150290 }, { "epoch": 3.003477079253427, "grad_norm": 0.11859495937824249, "learning_rate": 1.0016826412026724e-06, "loss": 0.0173, "step": 150300 }, { "epoch": 3.003676911394429, "grad_norm": 0.0910339504480362, "learning_rate": 9.996254987178266e-07, "loss": 0.0196, "step": 150310 }, { "epoch": 3.00387674353543, "grad_norm": 0.13211190700531006, "learning_rate": 9.97570449438967e-07, "loss": 0.0233, "step": 150320 }, { "epoch": 3.004076575676432, "grad_norm": 0.16378481686115265, "learning_rate": 9.955174934538736e-07, "loss": 0.0243, "step": 150330 }, { "epoch": 3.0042764078174335, "grad_norm": 0.08129724115133286, "learning_rate": 9.934666308502427e-07, "loss": 0.0332, "step": 150340 }, { "epoch": 3.004476239958435, "grad_norm": 0.1189679279923439, "learning_rate": 9.914178617156988e-07, "loss": 0.0246, "step": 150350 }, { "epoch": 3.0046760720994365, "grad_norm": 0.08588903397321701, "learning_rate": 9.893711861377442e-07, "loss": 0.0219, "step": 150360 }, { "epoch": 3.004875904240438, "grad_norm": 0.08155719190835953, "learning_rate": 9.873266042038198e-07, "loss": 0.0259, "step": 150370 }, { "epoch": 3.0050757363814395, "grad_norm": 0.16985338926315308, "learning_rate": 9.85284116001267e-07, "loss": 0.028, "step": 150380 }, { "epoch": 3.0052755685224413, "grad_norm": 0.20025496184825897, "learning_rate": 9.83243721617333e-07, "loss": 0.0361, "step": 150390 }, { "epoch": 3.0054754006634425, "grad_norm": 0.12219784408807755, "learning_rate": 9.812054211391752e-07, "loss": 0.0261, "step": 150400 }, { "epoch": 3.0056752328044443, "grad_norm": 0.06211962178349495, "learning_rate": 9.7916921465388e-07, "loss": 0.0165, "step": 150410 }, { "epoch": 3.005875064945446, "grad_norm": 0.1913551390171051, "learning_rate": 9.771351022484165e-07, "loss": 0.0243, "step": 150420 }, { "epoch": 3.0060748970864473, "grad_norm": 0.16394542157649994, "learning_rate": 9.75103084009682e-07, "loss": 0.023, "step": 150430 }, { "epoch": 3.006274729227449, "grad_norm": 0.10731583833694458, "learning_rate": 9.730731600244846e-07, "loss": 0.0283, "step": 150440 }, { "epoch": 3.0064745613684507, "grad_norm": 0.1412631720304489, "learning_rate": 9.710453303795385e-07, "loss": 0.0211, "step": 150450 }, { "epoch": 3.006674393509452, "grad_norm": 0.11420902609825134, "learning_rate": 9.690195951614688e-07, "loss": 0.0272, "step": 150460 }, { "epoch": 3.0068742256504537, "grad_norm": 0.15855076909065247, "learning_rate": 9.669959544568008e-07, "loss": 0.0196, "step": 150470 }, { "epoch": 3.007074057791455, "grad_norm": 0.10114312916994095, "learning_rate": 9.64974408351993e-07, "loss": 0.0259, "step": 150480 }, { "epoch": 3.0072738899324567, "grad_norm": 0.11689922958612442, "learning_rate": 9.629549569333983e-07, "loss": 0.0258, "step": 150490 }, { "epoch": 3.0074737220734584, "grad_norm": 0.11586841940879822, "learning_rate": 9.609376002872871e-07, "loss": 0.0215, "step": 150500 }, { "epoch": 3.0076735542144597, "grad_norm": 0.0885603055357933, "learning_rate": 9.589223384998292e-07, "loss": 0.0211, "step": 150510 }, { "epoch": 3.0078733863554614, "grad_norm": 0.13471204042434692, "learning_rate": 9.56909171657122e-07, "loss": 0.0282, "step": 150520 }, { "epoch": 3.008073218496463, "grad_norm": 0.10700605064630508, "learning_rate": 9.54898099845153e-07, "loss": 0.0209, "step": 150530 }, { "epoch": 3.0082730506374644, "grad_norm": 0.11230288445949554, "learning_rate": 9.528891231498472e-07, "loss": 0.0205, "step": 150540 }, { "epoch": 3.008472882778466, "grad_norm": 0.1296508014202118, "learning_rate": 9.508822416570084e-07, "loss": 0.0241, "step": 150550 }, { "epoch": 3.008672714919468, "grad_norm": 0.18102702498435974, "learning_rate": 9.488774554523794e-07, "loss": 0.0283, "step": 150560 }, { "epoch": 3.008872547060469, "grad_norm": 0.11592298001050949, "learning_rate": 9.468747646215915e-07, "loss": 0.0284, "step": 150570 }, { "epoch": 3.009072379201471, "grad_norm": 0.10358702391386032, "learning_rate": 9.448741692502038e-07, "loss": 0.0252, "step": 150580 }, { "epoch": 3.009272211342472, "grad_norm": 0.0993480235338211, "learning_rate": 9.428756694236762e-07, "loss": 0.0201, "step": 150590 }, { "epoch": 3.009472043483474, "grad_norm": 0.16601262986660004, "learning_rate": 9.408792652273734e-07, "loss": 0.0264, "step": 150600 }, { "epoch": 3.0096718756244756, "grad_norm": 0.11304228007793427, "learning_rate": 9.38884956746583e-07, "loss": 0.0258, "step": 150610 }, { "epoch": 3.009871707765477, "grad_norm": 0.10932699590921402, "learning_rate": 9.368927440665032e-07, "loss": 0.0222, "step": 150620 }, { "epoch": 3.0100715399064786, "grad_norm": 0.13116763532161713, "learning_rate": 9.349026272722327e-07, "loss": 0.0224, "step": 150630 }, { "epoch": 3.0102713720474803, "grad_norm": 0.10846754908561707, "learning_rate": 9.32914606448787e-07, "loss": 0.0228, "step": 150640 }, { "epoch": 3.0104712041884816, "grad_norm": 0.1257869452238083, "learning_rate": 9.309286816810924e-07, "loss": 0.0213, "step": 150650 }, { "epoch": 3.0106710363294833, "grad_norm": 0.15673108398914337, "learning_rate": 9.289448530539757e-07, "loss": 0.0254, "step": 150660 }, { "epoch": 3.0108708684704846, "grad_norm": 0.08885953575372696, "learning_rate": 9.269631206521911e-07, "loss": 0.029, "step": 150670 }, { "epoch": 3.0110707006114863, "grad_norm": 0.10800096392631531, "learning_rate": 9.249834845603933e-07, "loss": 0.0139, "step": 150680 }, { "epoch": 3.011270532752488, "grad_norm": 0.13173075020313263, "learning_rate": 9.230059448631534e-07, "loss": 0.0204, "step": 150690 }, { "epoch": 3.0114703648934893, "grad_norm": 0.18978251516819, "learning_rate": 9.210305016449317e-07, "loss": 0.029, "step": 150700 }, { "epoch": 3.011670197034491, "grad_norm": 0.09864308685064316, "learning_rate": 9.190571549901272e-07, "loss": 0.019, "step": 150710 }, { "epoch": 3.0118700291754927, "grad_norm": 0.11798958480358124, "learning_rate": 9.170859049830449e-07, "loss": 0.0188, "step": 150720 }, { "epoch": 3.012069861316494, "grad_norm": 0.07341080904006958, "learning_rate": 9.151167517078785e-07, "loss": 0.0159, "step": 150730 }, { "epoch": 3.0122696934574957, "grad_norm": 0.11249885708093643, "learning_rate": 9.131496952487495e-07, "loss": 0.0194, "step": 150740 }, { "epoch": 3.0124695255984975, "grad_norm": 0.10845624655485153, "learning_rate": 9.111847356896963e-07, "loss": 0.0261, "step": 150750 }, { "epoch": 3.0126693577394987, "grad_norm": 0.10717127472162247, "learning_rate": 9.092218731146463e-07, "loss": 0.038, "step": 150760 }, { "epoch": 3.0128691898805005, "grad_norm": 0.1668902486562729, "learning_rate": 9.0726110760746e-07, "loss": 0.0181, "step": 150770 }, { "epoch": 3.0130690220215017, "grad_norm": 0.12325847148895264, "learning_rate": 9.053024392518928e-07, "loss": 0.0231, "step": 150780 }, { "epoch": 3.0132688541625035, "grad_norm": 0.1856500506401062, "learning_rate": 9.03345868131611e-07, "loss": 0.0296, "step": 150790 }, { "epoch": 3.013468686303505, "grad_norm": 0.14891107380390167, "learning_rate": 9.013913943302033e-07, "loss": 0.0212, "step": 150800 }, { "epoch": 3.0136685184445065, "grad_norm": 0.10290220379829407, "learning_rate": 8.994390179311585e-07, "loss": 0.0254, "step": 150810 }, { "epoch": 3.013868350585508, "grad_norm": 0.1326225996017456, "learning_rate": 8.974887390178765e-07, "loss": 0.0256, "step": 150820 }, { "epoch": 3.01406818272651, "grad_norm": 0.19235484302043915, "learning_rate": 8.955405576736742e-07, "loss": 0.02, "step": 150830 }, { "epoch": 3.014268014867511, "grad_norm": 0.09114488214254379, "learning_rate": 8.935944739817681e-07, "loss": 0.0309, "step": 150840 }, { "epoch": 3.014467847008513, "grad_norm": 0.09981061518192291, "learning_rate": 8.916504880252974e-07, "loss": 0.0223, "step": 150850 }, { "epoch": 3.0146676791495146, "grad_norm": 0.12005093693733215, "learning_rate": 8.897085998873067e-07, "loss": 0.0223, "step": 150860 }, { "epoch": 3.014867511290516, "grad_norm": 0.2307186722755432, "learning_rate": 8.877688096507408e-07, "loss": 0.0214, "step": 150870 }, { "epoch": 3.0150673434315176, "grad_norm": 0.1254727691411972, "learning_rate": 8.858311173984668e-07, "loss": 0.0232, "step": 150880 }, { "epoch": 3.015267175572519, "grad_norm": 0.13093508780002594, "learning_rate": 8.838955232132684e-07, "loss": 0.0246, "step": 150890 }, { "epoch": 3.0154670077135206, "grad_norm": 0.09757853299379349, "learning_rate": 8.81962027177824e-07, "loss": 0.0246, "step": 150900 }, { "epoch": 3.0156668398545223, "grad_norm": 0.14687879383563995, "learning_rate": 8.800306293747229e-07, "loss": 0.0216, "step": 150910 }, { "epoch": 3.0158666719955236, "grad_norm": 0.15534597635269165, "learning_rate": 8.781013298864826e-07, "loss": 0.0229, "step": 150920 }, { "epoch": 3.0160665041365253, "grad_norm": 0.10056416690349579, "learning_rate": 8.761741287955095e-07, "loss": 0.0309, "step": 150930 }, { "epoch": 3.016266336277527, "grad_norm": 0.10679329186677933, "learning_rate": 8.742490261841374e-07, "loss": 0.0344, "step": 150940 }, { "epoch": 3.0164661684185283, "grad_norm": 0.14136433601379395, "learning_rate": 8.723260221345953e-07, "loss": 0.0215, "step": 150950 }, { "epoch": 3.01666600055953, "grad_norm": 0.13750267028808594, "learning_rate": 8.704051167290451e-07, "loss": 0.0374, "step": 150960 }, { "epoch": 3.0168658327005318, "grad_norm": 0.10170302540063858, "learning_rate": 8.684863100495211e-07, "loss": 0.0274, "step": 150970 }, { "epoch": 3.017065664841533, "grad_norm": 0.10931721329689026, "learning_rate": 8.665696021780134e-07, "loss": 0.025, "step": 150980 }, { "epoch": 3.0172654969825348, "grad_norm": 0.12515249848365784, "learning_rate": 8.646549931963898e-07, "loss": 0.0188, "step": 150990 }, { "epoch": 3.017465329123536, "grad_norm": 0.1197376474738121, "learning_rate": 8.627424831864406e-07, "loss": 0.0253, "step": 151000 }, { "epoch": 3.0176651612645378, "grad_norm": 0.10616623610258102, "learning_rate": 8.608320722298613e-07, "loss": 0.0246, "step": 151010 }, { "epoch": 3.0178649934055395, "grad_norm": 0.14317236840724945, "learning_rate": 8.5892376040827e-07, "loss": 0.0257, "step": 151020 }, { "epoch": 3.0180648255465408, "grad_norm": 0.13243752717971802, "learning_rate": 8.570175478031739e-07, "loss": 0.0247, "step": 151030 }, { "epoch": 3.0182646576875425, "grad_norm": 0.08830671012401581, "learning_rate": 8.551134344960132e-07, "loss": 0.0208, "step": 151040 }, { "epoch": 3.018464489828544, "grad_norm": 0.1305522322654724, "learning_rate": 8.53211420568123e-07, "loss": 0.022, "step": 151050 }, { "epoch": 3.0186643219695455, "grad_norm": 0.14300444722175598, "learning_rate": 8.513115061007604e-07, "loss": 0.0349, "step": 151060 }, { "epoch": 3.018864154110547, "grad_norm": 0.0909927487373352, "learning_rate": 8.494136911750716e-07, "loss": 0.02, "step": 151070 }, { "epoch": 3.0190639862515485, "grad_norm": 0.13870494067668915, "learning_rate": 8.475179758721474e-07, "loss": 0.0264, "step": 151080 }, { "epoch": 3.01926381839255, "grad_norm": 0.23836813867092133, "learning_rate": 8.456243602729563e-07, "loss": 0.0267, "step": 151090 }, { "epoch": 3.019463650533552, "grad_norm": 0.12531287968158722, "learning_rate": 8.437328444583891e-07, "loss": 0.0244, "step": 151100 }, { "epoch": 3.019663482674553, "grad_norm": 0.1315375417470932, "learning_rate": 8.418434285092536e-07, "loss": 0.0222, "step": 151110 }, { "epoch": 3.019863314815555, "grad_norm": 0.07877299189567566, "learning_rate": 8.399561125062627e-07, "loss": 0.021, "step": 151120 }, { "epoch": 3.0200631469565566, "grad_norm": 0.09243323653936386, "learning_rate": 8.380708965300355e-07, "loss": 0.0256, "step": 151130 }, { "epoch": 3.020262979097558, "grad_norm": 0.14057399332523346, "learning_rate": 8.361877806611074e-07, "loss": 0.0265, "step": 151140 }, { "epoch": 3.0204628112385596, "grad_norm": 0.08674784749746323, "learning_rate": 8.343067649799197e-07, "loss": 0.0264, "step": 151150 }, { "epoch": 3.0206626433795614, "grad_norm": 0.08304106444120407, "learning_rate": 8.324278495668247e-07, "loss": 0.0199, "step": 151160 }, { "epoch": 3.0208624755205626, "grad_norm": 0.10877688974142075, "learning_rate": 8.305510345020917e-07, "loss": 0.0232, "step": 151170 }, { "epoch": 3.0210623076615644, "grad_norm": 0.12769848108291626, "learning_rate": 8.286763198658898e-07, "loss": 0.0241, "step": 151180 }, { "epoch": 3.0212621398025656, "grad_norm": 0.08681413531303406, "learning_rate": 8.268037057383049e-07, "loss": 0.0178, "step": 151190 }, { "epoch": 3.0214619719435674, "grad_norm": 0.09074896574020386, "learning_rate": 8.249331921993286e-07, "loss": 0.0268, "step": 151200 }, { "epoch": 3.021661804084569, "grad_norm": 0.10155981034040451, "learning_rate": 8.230647793288693e-07, "loss": 0.0257, "step": 151210 }, { "epoch": 3.0218616362255704, "grad_norm": 0.15380869805812836, "learning_rate": 8.211984672067464e-07, "loss": 0.0257, "step": 151220 }, { "epoch": 3.022061468366572, "grad_norm": 0.10749328136444092, "learning_rate": 8.193342559126793e-07, "loss": 0.0234, "step": 151230 }, { "epoch": 3.022261300507574, "grad_norm": 0.12662842869758606, "learning_rate": 8.174721455263045e-07, "loss": 0.0218, "step": 151240 }, { "epoch": 3.022461132648575, "grad_norm": 0.12359455227851868, "learning_rate": 8.15612136127164e-07, "loss": 0.0271, "step": 151250 }, { "epoch": 3.022660964789577, "grad_norm": 0.1302628368139267, "learning_rate": 8.137542277947274e-07, "loss": 0.0244, "step": 151260 }, { "epoch": 3.0228607969305785, "grad_norm": 0.1530817747116089, "learning_rate": 8.11898420608348e-07, "loss": 0.0201, "step": 151270 }, { "epoch": 3.02306062907158, "grad_norm": 0.07976885885000229, "learning_rate": 8.100447146473012e-07, "loss": 0.0243, "step": 151280 }, { "epoch": 3.0232604612125815, "grad_norm": 0.14317335188388824, "learning_rate": 8.081931099907846e-07, "loss": 0.022, "step": 151290 }, { "epoch": 3.023460293353583, "grad_norm": 0.14292529225349426, "learning_rate": 8.063436067178909e-07, "loss": 0.0258, "step": 151300 }, { "epoch": 3.0236601254945845, "grad_norm": 0.08677882701158524, "learning_rate": 8.044962049076232e-07, "loss": 0.0256, "step": 151310 }, { "epoch": 3.0238599576355862, "grad_norm": 0.09736865013837814, "learning_rate": 8.026509046389075e-07, "loss": 0.0268, "step": 151320 }, { "epoch": 3.0240597897765875, "grad_norm": 0.1330118179321289, "learning_rate": 8.008077059905639e-07, "loss": 0.0223, "step": 151330 }, { "epoch": 3.0242596219175892, "grad_norm": 0.1541576236486435, "learning_rate": 7.989666090413295e-07, "loss": 0.0215, "step": 151340 }, { "epoch": 3.024459454058591, "grad_norm": 0.1079266220331192, "learning_rate": 7.971276138698635e-07, "loss": 0.0283, "step": 151350 }, { "epoch": 3.0246592861995922, "grad_norm": 0.10988776385784149, "learning_rate": 7.952907205547144e-07, "loss": 0.0232, "step": 151360 }, { "epoch": 3.024859118340594, "grad_norm": 0.13899537920951843, "learning_rate": 7.934559291743526e-07, "loss": 0.0252, "step": 151370 }, { "epoch": 3.0250589504815957, "grad_norm": 0.10714145749807358, "learning_rate": 7.916232398071599e-07, "loss": 0.0233, "step": 151380 }, { "epoch": 3.025258782622597, "grad_norm": 0.09226511418819427, "learning_rate": 7.897926525314237e-07, "loss": 0.0239, "step": 151390 }, { "epoch": 3.0254586147635987, "grad_norm": 0.09355148673057556, "learning_rate": 7.879641674253424e-07, "loss": 0.0212, "step": 151400 }, { "epoch": 3.0256584469046, "grad_norm": 0.16156110167503357, "learning_rate": 7.861377845670259e-07, "loss": 0.0201, "step": 151410 }, { "epoch": 3.0258582790456017, "grad_norm": 0.13686470687389374, "learning_rate": 7.843135040344951e-07, "loss": 0.0209, "step": 151420 }, { "epoch": 3.0260581111866034, "grad_norm": 0.08726069331169128, "learning_rate": 7.824913259056765e-07, "loss": 0.0221, "step": 151430 }, { "epoch": 3.0262579433276047, "grad_norm": 0.10001611709594727, "learning_rate": 7.806712502584191e-07, "loss": 0.0285, "step": 151440 }, { "epoch": 3.0264577754686064, "grad_norm": 0.0699385330080986, "learning_rate": 7.788532771704604e-07, "loss": 0.0205, "step": 151450 }, { "epoch": 3.026657607609608, "grad_norm": 0.12340715527534485, "learning_rate": 7.770374067194719e-07, "loss": 0.0364, "step": 151460 }, { "epoch": 3.0268574397506094, "grad_norm": 0.0999220684170723, "learning_rate": 7.752236389830192e-07, "loss": 0.019, "step": 151470 }, { "epoch": 3.027057271891611, "grad_norm": 0.08996263146400452, "learning_rate": 7.734119740385848e-07, "loss": 0.0258, "step": 151480 }, { "epoch": 3.0272571040326124, "grad_norm": 0.10227320343255997, "learning_rate": 7.716024119635567e-07, "loss": 0.0272, "step": 151490 }, { "epoch": 3.027456936173614, "grad_norm": 0.09722588956356049, "learning_rate": 7.6979495283524e-07, "loss": 0.0228, "step": 151500 }, { "epoch": 3.027656768314616, "grad_norm": 0.5438796281814575, "learning_rate": 7.679895967308392e-07, "loss": 0.0242, "step": 151510 }, { "epoch": 3.027856600455617, "grad_norm": 0.10589318722486496, "learning_rate": 7.661863437274875e-07, "loss": 0.0304, "step": 151520 }, { "epoch": 3.028056432596619, "grad_norm": 0.09771689772605896, "learning_rate": 7.643851939022062e-07, "loss": 0.0159, "step": 151530 }, { "epoch": 3.0282562647376206, "grad_norm": 0.08722522854804993, "learning_rate": 7.625861473319451e-07, "loss": 0.0187, "step": 151540 }, { "epoch": 3.028456096878622, "grad_norm": 0.11807796359062195, "learning_rate": 7.607892040935427e-07, "loss": 0.0245, "step": 151550 }, { "epoch": 3.0286559290196235, "grad_norm": 0.10917919874191284, "learning_rate": 7.589943642637765e-07, "loss": 0.0188, "step": 151560 }, { "epoch": 3.0288557611606253, "grad_norm": 0.1082104966044426, "learning_rate": 7.572016279193128e-07, "loss": 0.0254, "step": 151570 }, { "epoch": 3.0290555933016265, "grad_norm": 0.12264884263277054, "learning_rate": 7.554109951367294e-07, "loss": 0.0242, "step": 151580 }, { "epoch": 3.0292554254426283, "grad_norm": 0.11232196539640427, "learning_rate": 7.536224659925261e-07, "loss": 0.0264, "step": 151590 }, { "epoch": 3.0294552575836295, "grad_norm": 0.14525073766708374, "learning_rate": 7.518360405631031e-07, "loss": 0.0227, "step": 151600 }, { "epoch": 3.0296550897246313, "grad_norm": 0.13655583560466766, "learning_rate": 7.500517189247713e-07, "loss": 0.0158, "step": 151610 }, { "epoch": 3.029854921865633, "grad_norm": 0.10743070393800735, "learning_rate": 7.482695011537588e-07, "loss": 0.0246, "step": 151620 }, { "epoch": 3.0300547540066343, "grad_norm": 0.13453052937984467, "learning_rate": 7.464893873261936e-07, "loss": 0.0197, "step": 151630 }, { "epoch": 3.030254586147636, "grad_norm": 0.10906119644641876, "learning_rate": 7.447113775181147e-07, "loss": 0.0261, "step": 151640 }, { "epoch": 3.0304544182886377, "grad_norm": 0.10790936648845673, "learning_rate": 7.429354718054837e-07, "loss": 0.0312, "step": 151650 }, { "epoch": 3.030654250429639, "grad_norm": 0.14153248071670532, "learning_rate": 7.411616702641677e-07, "loss": 0.0199, "step": 151660 }, { "epoch": 3.0308540825706407, "grad_norm": 0.06834287941455841, "learning_rate": 7.393899729699227e-07, "loss": 0.0272, "step": 151670 }, { "epoch": 3.0310539147116424, "grad_norm": 0.11354594677686691, "learning_rate": 7.376203799984494e-07, "loss": 0.0329, "step": 151680 }, { "epoch": 3.0312537468526437, "grad_norm": 0.10784194618463516, "learning_rate": 7.358528914253371e-07, "loss": 0.0274, "step": 151690 }, { "epoch": 3.0314535789936454, "grad_norm": 0.0993431806564331, "learning_rate": 7.340875073260867e-07, "loss": 0.0267, "step": 151700 }, { "epoch": 3.0316534111346467, "grad_norm": 0.14954930543899536, "learning_rate": 7.3232422777611e-07, "loss": 0.0321, "step": 151710 }, { "epoch": 3.0318532432756484, "grad_norm": 0.09538895636796951, "learning_rate": 7.305630528507412e-07, "loss": 0.0208, "step": 151720 }, { "epoch": 3.03205307541665, "grad_norm": 0.10052444040775299, "learning_rate": 7.288039826252035e-07, "loss": 0.0305, "step": 151730 }, { "epoch": 3.0322529075576514, "grad_norm": 0.09699925035238266, "learning_rate": 7.270470171746424e-07, "loss": 0.0244, "step": 151740 }, { "epoch": 3.032452739698653, "grad_norm": 0.09205742925405502, "learning_rate": 7.2529215657412e-07, "loss": 0.0156, "step": 151750 }, { "epoch": 3.032652571839655, "grad_norm": 0.14394523203372955, "learning_rate": 7.235394008985985e-07, "loss": 0.0202, "step": 151760 }, { "epoch": 3.032852403980656, "grad_norm": 0.15106044709682465, "learning_rate": 7.217887502229459e-07, "loss": 0.0198, "step": 151770 }, { "epoch": 3.033052236121658, "grad_norm": 0.10580039024353027, "learning_rate": 7.200402046219523e-07, "loss": 0.0243, "step": 151780 }, { "epoch": 3.0332520682626596, "grad_norm": 0.10880695283412933, "learning_rate": 7.182937641703136e-07, "loss": 0.0248, "step": 151790 }, { "epoch": 3.033451900403661, "grad_norm": 0.1428183764219284, "learning_rate": 7.165494289426311e-07, "loss": 0.0226, "step": 151800 }, { "epoch": 3.0336517325446626, "grad_norm": 0.07483838498592377, "learning_rate": 7.148071990134175e-07, "loss": 0.0341, "step": 151810 }, { "epoch": 3.033851564685664, "grad_norm": 0.10213501006364822, "learning_rate": 7.130670744571022e-07, "loss": 0.0173, "step": 151820 }, { "epoch": 3.0340513968266656, "grad_norm": 0.10952670872211456, "learning_rate": 7.113290553480145e-07, "loss": 0.024, "step": 151830 }, { "epoch": 3.0342512289676673, "grad_norm": 0.1067308634519577, "learning_rate": 7.095931417604063e-07, "loss": 0.0208, "step": 151840 }, { "epoch": 3.0344510611086686, "grad_norm": 0.1124083548784256, "learning_rate": 7.078593337684291e-07, "loss": 0.021, "step": 151850 }, { "epoch": 3.0346508932496703, "grad_norm": 0.13876977562904358, "learning_rate": 7.061276314461518e-07, "loss": 0.0194, "step": 151860 }, { "epoch": 3.034850725390672, "grad_norm": 0.09683909267187119, "learning_rate": 7.043980348675427e-07, "loss": 0.0312, "step": 151870 }, { "epoch": 3.0350505575316733, "grad_norm": 0.08497685194015503, "learning_rate": 7.026705441064929e-07, "loss": 0.0282, "step": 151880 }, { "epoch": 3.035250389672675, "grad_norm": 0.07957493513822556, "learning_rate": 7.009451592367988e-07, "loss": 0.021, "step": 151890 }, { "epoch": 3.0354502218136763, "grad_norm": 0.11440369486808777, "learning_rate": 6.992218803321571e-07, "loss": 0.023, "step": 151900 }, { "epoch": 3.035650053954678, "grad_norm": 0.12244876474142075, "learning_rate": 6.975007074661865e-07, "loss": 0.0288, "step": 151910 }, { "epoch": 3.0358498860956797, "grad_norm": 0.10080213099718094, "learning_rate": 6.957816407124174e-07, "loss": 0.0195, "step": 151920 }, { "epoch": 3.036049718236681, "grad_norm": 0.134116530418396, "learning_rate": 6.940646801442851e-07, "loss": 0.0208, "step": 151930 }, { "epoch": 3.0362495503776827, "grad_norm": 0.10483177751302719, "learning_rate": 6.923498258351257e-07, "loss": 0.0231, "step": 151940 }, { "epoch": 3.0364493825186845, "grad_norm": 0.1030726507306099, "learning_rate": 6.906370778581972e-07, "loss": 0.0213, "step": 151950 }, { "epoch": 3.0366492146596857, "grad_norm": 0.12787874042987823, "learning_rate": 6.889264362866798e-07, "loss": 0.0227, "step": 151960 }, { "epoch": 3.0368490468006875, "grad_norm": 0.11314716190099716, "learning_rate": 6.872179011936319e-07, "loss": 0.0193, "step": 151970 }, { "epoch": 3.037048878941689, "grad_norm": 0.10305105894804001, "learning_rate": 6.855114726520451e-07, "loss": 0.0247, "step": 151980 }, { "epoch": 3.0372487110826905, "grad_norm": 0.10982853174209595, "learning_rate": 6.838071507348109e-07, "loss": 0.0274, "step": 151990 }, { "epoch": 3.037448543223692, "grad_norm": 0.11189384758472443, "learning_rate": 6.821049355147435e-07, "loss": 0.0208, "step": 152000 }, { "epoch": 3.0376483753646935, "grad_norm": 0.11349613219499588, "learning_rate": 6.804048270645513e-07, "loss": 0.0156, "step": 152010 }, { "epoch": 3.037848207505695, "grad_norm": 0.13413579761981964, "learning_rate": 6.787068254568652e-07, "loss": 0.0152, "step": 152020 }, { "epoch": 3.038048039646697, "grad_norm": 0.10069132596254349, "learning_rate": 6.770109307642159e-07, "loss": 0.0267, "step": 152030 }, { "epoch": 3.038247871787698, "grad_norm": 0.10310754179954529, "learning_rate": 6.753171430590454e-07, "loss": 0.0222, "step": 152040 }, { "epoch": 3.0384477039287, "grad_norm": 0.09079157561063766, "learning_rate": 6.736254624137183e-07, "loss": 0.0202, "step": 152050 }, { "epoch": 3.0386475360697016, "grad_norm": 0.07700233906507492, "learning_rate": 6.71935888900499e-07, "loss": 0.02, "step": 152060 }, { "epoch": 3.038847368210703, "grad_norm": 0.11186766624450684, "learning_rate": 6.702484225915573e-07, "loss": 0.0276, "step": 152070 }, { "epoch": 3.0390472003517046, "grad_norm": 0.10503140836954117, "learning_rate": 6.685630635589801e-07, "loss": 0.0269, "step": 152080 }, { "epoch": 3.0392470324927063, "grad_norm": 0.13354407250881195, "learning_rate": 6.668798118747655e-07, "loss": 0.0181, "step": 152090 }, { "epoch": 3.0394468646337076, "grad_norm": 0.10579651594161987, "learning_rate": 6.651986676108169e-07, "loss": 0.0272, "step": 152100 }, { "epoch": 3.0396466967747093, "grad_norm": 0.14856795966625214, "learning_rate": 6.635196308389491e-07, "loss": 0.0234, "step": 152110 }, { "epoch": 3.0398465289157106, "grad_norm": 0.10261990875005722, "learning_rate": 6.618427016308881e-07, "loss": 0.0125, "step": 152120 }, { "epoch": 3.0400463610567123, "grad_norm": 0.12800724804401398, "learning_rate": 6.601678800582711e-07, "loss": 0.0266, "step": 152130 }, { "epoch": 3.040246193197714, "grad_norm": 0.12824313342571259, "learning_rate": 6.584951661926409e-07, "loss": 0.0217, "step": 152140 }, { "epoch": 3.0404460253387153, "grad_norm": 0.12395475059747696, "learning_rate": 6.568245601054568e-07, "loss": 0.0216, "step": 152150 }, { "epoch": 3.040645857479717, "grad_norm": 0.11499106138944626, "learning_rate": 6.551560618680841e-07, "loss": 0.0248, "step": 152160 }, { "epoch": 3.0408456896207188, "grad_norm": 0.12788482010364532, "learning_rate": 6.53489671551788e-07, "loss": 0.0228, "step": 152170 }, { "epoch": 3.04104552176172, "grad_norm": 0.09450917690992355, "learning_rate": 6.518253892277615e-07, "loss": 0.0251, "step": 152180 }, { "epoch": 3.0412453539027218, "grad_norm": 0.11339259147644043, "learning_rate": 6.501632149671033e-07, "loss": 0.0207, "step": 152190 }, { "epoch": 3.041445186043723, "grad_norm": 0.12905776500701904, "learning_rate": 6.485031488408122e-07, "loss": 0.0233, "step": 152200 }, { "epoch": 3.0416450181847248, "grad_norm": 0.09954459965229034, "learning_rate": 6.468451909198037e-07, "loss": 0.0289, "step": 152210 }, { "epoch": 3.0418448503257265, "grad_norm": 0.19329169392585754, "learning_rate": 6.451893412749044e-07, "loss": 0.0191, "step": 152220 }, { "epoch": 3.0420446824667278, "grad_norm": 0.07569152116775513, "learning_rate": 6.435355999768522e-07, "loss": 0.035, "step": 152230 }, { "epoch": 3.0422445146077295, "grad_norm": 0.1155388206243515, "learning_rate": 6.418839670962851e-07, "loss": 0.0217, "step": 152240 }, { "epoch": 3.042444346748731, "grad_norm": 0.1092401072382927, "learning_rate": 6.402344427037577e-07, "loss": 0.0238, "step": 152250 }, { "epoch": 3.0426441788897325, "grad_norm": 0.10853961110115051, "learning_rate": 6.385870268697469e-07, "loss": 0.029, "step": 152260 }, { "epoch": 3.042844011030734, "grad_norm": 0.09723832458257675, "learning_rate": 6.369417196646077e-07, "loss": 0.0263, "step": 152270 }, { "epoch": 3.043043843171736, "grad_norm": 0.178677037358284, "learning_rate": 6.352985211586448e-07, "loss": 0.0228, "step": 152280 }, { "epoch": 3.043243675312737, "grad_norm": 0.16341739892959595, "learning_rate": 6.336574314220412e-07, "loss": 0.028, "step": 152290 }, { "epoch": 3.043443507453739, "grad_norm": 0.09165194630622864, "learning_rate": 6.320184505249016e-07, "loss": 0.0307, "step": 152300 }, { "epoch": 3.04364333959474, "grad_norm": 0.12809927761554718, "learning_rate": 6.303815785372369e-07, "loss": 0.0249, "step": 152310 }, { "epoch": 3.043843171735742, "grad_norm": 0.12282268702983856, "learning_rate": 6.287468155289855e-07, "loss": 0.0252, "step": 152320 }, { "epoch": 3.0440430038767436, "grad_norm": 0.1952013522386551, "learning_rate": 6.271141615699693e-07, "loss": 0.0243, "step": 152330 }, { "epoch": 3.044242836017745, "grad_norm": 0.1534382551908493, "learning_rate": 6.254836167299327e-07, "loss": 0.0236, "step": 152340 }, { "epoch": 3.0444426681587466, "grad_norm": 0.09503937512636185, "learning_rate": 6.238551810785309e-07, "loss": 0.0238, "step": 152350 }, { "epoch": 3.0446425002997484, "grad_norm": 0.07504480332136154, "learning_rate": 6.222288546853306e-07, "loss": 0.0266, "step": 152360 }, { "epoch": 3.0448423324407496, "grad_norm": 0.14437046647071838, "learning_rate": 6.206046376198038e-07, "loss": 0.0199, "step": 152370 }, { "epoch": 3.0450421645817514, "grad_norm": 0.144766703248024, "learning_rate": 6.189825299513341e-07, "loss": 0.0182, "step": 152380 }, { "epoch": 3.045241996722753, "grad_norm": 0.10185346752405167, "learning_rate": 6.17362531749216e-07, "loss": 0.0253, "step": 152390 }, { "epoch": 3.0454418288637544, "grad_norm": 0.08320309966802597, "learning_rate": 6.157446430826552e-07, "loss": 0.0169, "step": 152400 }, { "epoch": 3.045641661004756, "grad_norm": 0.1700129359960556, "learning_rate": 6.141288640207577e-07, "loss": 0.021, "step": 152410 }, { "epoch": 3.0458414931457574, "grad_norm": 0.12896984815597534, "learning_rate": 6.125151946325458e-07, "loss": 0.0269, "step": 152420 }, { "epoch": 3.046041325286759, "grad_norm": 0.12492246925830841, "learning_rate": 6.109036349869646e-07, "loss": 0.0218, "step": 152430 }, { "epoch": 3.046241157427761, "grad_norm": 0.14907075464725494, "learning_rate": 6.092941851528477e-07, "loss": 0.0247, "step": 152440 }, { "epoch": 3.046440989568762, "grad_norm": 0.09236367791891098, "learning_rate": 6.07686845198946e-07, "loss": 0.0251, "step": 152450 }, { "epoch": 3.046640821709764, "grad_norm": 0.10024955123662949, "learning_rate": 6.060816151939319e-07, "loss": 0.0219, "step": 152460 }, { "epoch": 3.0468406538507655, "grad_norm": 0.10395515710115433, "learning_rate": 6.044784952063675e-07, "loss": 0.0309, "step": 152470 }, { "epoch": 3.047040485991767, "grad_norm": 0.14229480922222137, "learning_rate": 6.02877485304737e-07, "loss": 0.0161, "step": 152480 }, { "epoch": 3.0472403181327685, "grad_norm": 0.12044627964496613, "learning_rate": 6.012785855574355e-07, "loss": 0.0285, "step": 152490 }, { "epoch": 3.0474401502737702, "grad_norm": 0.09675858169794083, "learning_rate": 5.996817960327695e-07, "loss": 0.0204, "step": 152500 }, { "epoch": 3.0476399824147715, "grad_norm": 0.1428166627883911, "learning_rate": 5.980871167989399e-07, "loss": 0.0267, "step": 152510 }, { "epoch": 3.0478398145557732, "grad_norm": 0.15911823511123657, "learning_rate": 5.964945479240757e-07, "loss": 0.0211, "step": 152520 }, { "epoch": 3.0480396466967745, "grad_norm": 0.09976239502429962, "learning_rate": 5.94904089476217e-07, "loss": 0.0205, "step": 152530 }, { "epoch": 3.0482394788377762, "grad_norm": 0.10026448220014572, "learning_rate": 5.933157415232815e-07, "loss": 0.0178, "step": 152540 }, { "epoch": 3.048439310978778, "grad_norm": 0.10525037348270416, "learning_rate": 5.917295041331427e-07, "loss": 0.0203, "step": 152550 }, { "epoch": 3.0486391431197792, "grad_norm": 0.12083345651626587, "learning_rate": 5.901453773735522e-07, "loss": 0.0254, "step": 152560 }, { "epoch": 3.048838975260781, "grad_norm": 0.11253079026937485, "learning_rate": 5.885633613121832e-07, "loss": 0.0267, "step": 152570 }, { "epoch": 3.0490388074017827, "grad_norm": 0.153070330619812, "learning_rate": 5.869834560166099e-07, "loss": 0.0325, "step": 152580 }, { "epoch": 3.049238639542784, "grad_norm": 0.13443607091903687, "learning_rate": 5.854056615543335e-07, "loss": 0.0301, "step": 152590 }, { "epoch": 3.0494384716837857, "grad_norm": 0.19497919082641602, "learning_rate": 5.838299779927503e-07, "loss": 0.0251, "step": 152600 }, { "epoch": 3.0496383038247874, "grad_norm": 0.23108193278312683, "learning_rate": 5.822564053991675e-07, "loss": 0.0255, "step": 152610 }, { "epoch": 3.0498381359657887, "grad_norm": 0.11018295586109161, "learning_rate": 5.806849438408091e-07, "loss": 0.0216, "step": 152620 }, { "epoch": 3.0500379681067904, "grad_norm": 0.15318258106708527, "learning_rate": 5.791155933848046e-07, "loss": 0.0266, "step": 152630 }, { "epoch": 3.0502378002477917, "grad_norm": 0.11625208705663681, "learning_rate": 5.775483540981896e-07, "loss": 0.0232, "step": 152640 }, { "epoch": 3.0504376323887934, "grad_norm": 0.09930624067783356, "learning_rate": 5.759832260479214e-07, "loss": 0.0244, "step": 152650 }, { "epoch": 3.050637464529795, "grad_norm": 0.11898288875818253, "learning_rate": 5.744202093008577e-07, "loss": 0.0166, "step": 152660 }, { "epoch": 3.0508372966707964, "grad_norm": 0.10045792162418365, "learning_rate": 5.728593039237562e-07, "loss": 0.0277, "step": 152670 }, { "epoch": 3.051037128811798, "grad_norm": 0.12711583077907562, "learning_rate": 5.713005099833079e-07, "loss": 0.0268, "step": 152680 }, { "epoch": 3.0512369609528, "grad_norm": 0.0937371626496315, "learning_rate": 5.697438275460986e-07, "loss": 0.0293, "step": 152690 }, { "epoch": 3.051436793093801, "grad_norm": 0.08841124176979065, "learning_rate": 5.681892566786306e-07, "loss": 0.0225, "step": 152700 }, { "epoch": 3.051636625234803, "grad_norm": 0.18122906982898712, "learning_rate": 5.666367974473063e-07, "loss": 0.0268, "step": 152710 }, { "epoch": 3.051836457375804, "grad_norm": 0.1380765438079834, "learning_rate": 5.65086449918445e-07, "loss": 0.0215, "step": 152720 }, { "epoch": 3.052036289516806, "grad_norm": 0.09380769729614258, "learning_rate": 5.635382141582712e-07, "loss": 0.0283, "step": 152730 }, { "epoch": 3.0522361216578076, "grad_norm": 0.11562322825193405, "learning_rate": 5.619920902329323e-07, "loss": 0.0255, "step": 152740 }, { "epoch": 3.052435953798809, "grad_norm": 0.11756082624197006, "learning_rate": 5.604480782084698e-07, "loss": 0.0259, "step": 152750 }, { "epoch": 3.0526357859398106, "grad_norm": 0.11038399487733841, "learning_rate": 5.58906178150842e-07, "loss": 0.0243, "step": 152760 }, { "epoch": 3.0528356180808123, "grad_norm": 0.07713941484689713, "learning_rate": 5.573663901259185e-07, "loss": 0.0237, "step": 152770 }, { "epoch": 3.0530354502218136, "grad_norm": 0.11838429421186447, "learning_rate": 5.558287141994689e-07, "loss": 0.0157, "step": 152780 }, { "epoch": 3.0532352823628153, "grad_norm": 0.12540508806705475, "learning_rate": 5.542931504371906e-07, "loss": 0.0231, "step": 152790 }, { "epoch": 3.053435114503817, "grad_norm": 0.11634540557861328, "learning_rate": 5.527596989046702e-07, "loss": 0.024, "step": 152800 }, { "epoch": 3.0536349466448183, "grad_norm": 0.10146190226078033, "learning_rate": 5.512283596674162e-07, "loss": 0.0317, "step": 152810 }, { "epoch": 3.05383477878582, "grad_norm": 0.13119672238826752, "learning_rate": 5.496991327908485e-07, "loss": 0.0203, "step": 152820 }, { "epoch": 3.0540346109268213, "grad_norm": 0.11601501703262329, "learning_rate": 5.481720183402983e-07, "loss": 0.0293, "step": 152830 }, { "epoch": 3.054234443067823, "grad_norm": 0.08630086481571198, "learning_rate": 5.466470163809855e-07, "loss": 0.0283, "step": 152840 }, { "epoch": 3.0544342752088247, "grad_norm": 0.09533103555440903, "learning_rate": 5.451241269780583e-07, "loss": 0.0215, "step": 152850 }, { "epoch": 3.054634107349826, "grad_norm": 0.16628260910511017, "learning_rate": 5.436033501965864e-07, "loss": 0.0189, "step": 152860 }, { "epoch": 3.0548339394908277, "grad_norm": 0.09441042691469193, "learning_rate": 5.420846861015239e-07, "loss": 0.0211, "step": 152870 }, { "epoch": 3.0550337716318294, "grad_norm": 0.09405460953712463, "learning_rate": 5.405681347577407e-07, "loss": 0.0231, "step": 152880 }, { "epoch": 3.0552336037728307, "grad_norm": 0.10910306870937347, "learning_rate": 5.390536962300352e-07, "loss": 0.0211, "step": 152890 }, { "epoch": 3.0554334359138324, "grad_norm": 0.1254691779613495, "learning_rate": 5.375413705830834e-07, "loss": 0.0288, "step": 152900 }, { "epoch": 3.055633268054834, "grad_norm": 0.12139906734228134, "learning_rate": 5.360311578815058e-07, "loss": 0.0174, "step": 152910 }, { "epoch": 3.0558331001958354, "grad_norm": 0.09490521997213364, "learning_rate": 5.345230581898065e-07, "loss": 0.0218, "step": 152920 }, { "epoch": 3.056032932336837, "grad_norm": 0.10474585741758347, "learning_rate": 5.330170715724169e-07, "loss": 0.02, "step": 152930 }, { "epoch": 3.0562327644778384, "grad_norm": 0.09517651051282883, "learning_rate": 5.315131980936528e-07, "loss": 0.0159, "step": 152940 }, { "epoch": 3.05643259661884, "grad_norm": 0.09860903024673462, "learning_rate": 5.30011437817779e-07, "loss": 0.0198, "step": 152950 }, { "epoch": 3.056632428759842, "grad_norm": 0.11757183074951172, "learning_rate": 5.285117908089332e-07, "loss": 0.0247, "step": 152960 }, { "epoch": 3.056832260900843, "grad_norm": 0.1894940882921219, "learning_rate": 5.270142571311809e-07, "loss": 0.0214, "step": 152970 }, { "epoch": 3.057032093041845, "grad_norm": 0.0728580430150032, "learning_rate": 5.255188368484931e-07, "loss": 0.0197, "step": 152980 }, { "epoch": 3.0572319251828466, "grad_norm": 0.16205467283725739, "learning_rate": 5.240255300247577e-07, "loss": 0.0287, "step": 152990 }, { "epoch": 3.057431757323848, "grad_norm": 0.09996864199638367, "learning_rate": 5.225343367237567e-07, "loss": 0.0195, "step": 153000 }, { "epoch": 3.0576315894648496, "grad_norm": 0.11671458184719086, "learning_rate": 5.210452570092006e-07, "loss": 0.024, "step": 153010 }, { "epoch": 3.057831421605851, "grad_norm": 0.12293004244565964, "learning_rate": 5.195582909446939e-07, "loss": 0.0182, "step": 153020 }, { "epoch": 3.0580312537468526, "grad_norm": 0.11726079881191254, "learning_rate": 5.18073438593758e-07, "loss": 0.0225, "step": 153030 }, { "epoch": 3.0582310858878543, "grad_norm": 0.12224402278661728, "learning_rate": 5.1659070001982e-07, "loss": 0.028, "step": 153040 }, { "epoch": 3.0584309180288556, "grad_norm": 0.09899149090051651, "learning_rate": 5.151100752862292e-07, "loss": 0.0321, "step": 153050 }, { "epoch": 3.0586307501698573, "grad_norm": 0.12145013362169266, "learning_rate": 5.136315644562295e-07, "loss": 0.0243, "step": 153060 }, { "epoch": 3.058830582310859, "grad_norm": 0.13325844705104828, "learning_rate": 5.121551675929703e-07, "loss": 0.0268, "step": 153070 }, { "epoch": 3.0590304144518603, "grad_norm": 0.13316698372364044, "learning_rate": 5.106808847595401e-07, "loss": 0.0201, "step": 153080 }, { "epoch": 3.059230246592862, "grad_norm": 0.09131447225809097, "learning_rate": 5.092087160189051e-07, "loss": 0.0267, "step": 153090 }, { "epoch": 3.0594300787338637, "grad_norm": 0.10428030788898468, "learning_rate": 5.077386614339541e-07, "loss": 0.0233, "step": 153100 }, { "epoch": 3.059629910874865, "grad_norm": 0.10902005434036255, "learning_rate": 5.062707210674921e-07, "loss": 0.0187, "step": 153110 }, { "epoch": 3.0598297430158667, "grad_norm": 0.11877177655696869, "learning_rate": 5.048048949822193e-07, "loss": 0.0229, "step": 153120 }, { "epoch": 3.060029575156868, "grad_norm": 0.0871957540512085, "learning_rate": 5.03341183240752e-07, "loss": 0.0221, "step": 153130 }, { "epoch": 3.0602294072978697, "grad_norm": 0.11496451497077942, "learning_rate": 5.018795859056236e-07, "loss": 0.0215, "step": 153140 }, { "epoch": 3.0604292394388715, "grad_norm": 0.14232148230075836, "learning_rate": 5.00420103039273e-07, "loss": 0.0251, "step": 153150 }, { "epoch": 3.0606290715798727, "grad_norm": 0.11583767086267471, "learning_rate": 4.989627347040338e-07, "loss": 0.0275, "step": 153160 }, { "epoch": 3.0608289037208745, "grad_norm": 0.1182972863316536, "learning_rate": 4.975074809621782e-07, "loss": 0.0322, "step": 153170 }, { "epoch": 3.061028735861876, "grad_norm": 0.11400771141052246, "learning_rate": 4.960543418758567e-07, "loss": 0.0181, "step": 153180 }, { "epoch": 3.0612285680028775, "grad_norm": 0.09019825607538223, "learning_rate": 4.946033175071585e-07, "loss": 0.0291, "step": 153190 }, { "epoch": 3.061428400143879, "grad_norm": 0.11275625228881836, "learning_rate": 4.931544079180616e-07, "loss": 0.0283, "step": 153200 }, { "epoch": 3.061628232284881, "grad_norm": 0.15869547426700592, "learning_rate": 4.917076131704556e-07, "loss": 0.0182, "step": 153210 }, { "epoch": 3.061828064425882, "grad_norm": 0.13109852373600006, "learning_rate": 4.902629333261577e-07, "loss": 0.0224, "step": 153220 }, { "epoch": 3.062027896566884, "grad_norm": 0.0837821215391159, "learning_rate": 4.88820368446874e-07, "loss": 0.0284, "step": 153230 }, { "epoch": 3.062227728707885, "grad_norm": 0.12416121363639832, "learning_rate": 4.873799185942274e-07, "loss": 0.0249, "step": 153240 }, { "epoch": 3.062427560848887, "grad_norm": 0.11048656702041626, "learning_rate": 4.859415838297465e-07, "loss": 0.0244, "step": 153250 }, { "epoch": 3.0626273929898886, "grad_norm": 0.07600117474794388, "learning_rate": 4.845053642148934e-07, "loss": 0.015, "step": 153260 }, { "epoch": 3.06282722513089, "grad_norm": 0.15792441368103027, "learning_rate": 4.830712598110021e-07, "loss": 0.0205, "step": 153270 }, { "epoch": 3.0630270572718916, "grad_norm": 0.12396909296512604, "learning_rate": 4.816392706793405e-07, "loss": 0.0338, "step": 153280 }, { "epoch": 3.0632268894128933, "grad_norm": 0.13471198081970215, "learning_rate": 4.802093968810817e-07, "loss": 0.0193, "step": 153290 }, { "epoch": 3.0634267215538946, "grad_norm": 0.09778999537229538, "learning_rate": 4.787816384773047e-07, "loss": 0.0241, "step": 153300 }, { "epoch": 3.0636265536948963, "grad_norm": 0.1608150601387024, "learning_rate": 4.773559955290052e-07, "loss": 0.0229, "step": 153310 }, { "epoch": 3.063826385835898, "grad_norm": 0.11390021443367004, "learning_rate": 4.7593246809707893e-07, "loss": 0.023, "step": 153320 }, { "epoch": 3.0640262179768993, "grad_norm": 0.14014960825443268, "learning_rate": 4.745110562423438e-07, "loss": 0.0254, "step": 153330 }, { "epoch": 3.064226050117901, "grad_norm": 0.09821755439043045, "learning_rate": 4.7309176002551246e-07, "loss": 0.0233, "step": 153340 }, { "epoch": 3.0644258822589023, "grad_norm": 0.148333340883255, "learning_rate": 4.7167457950721416e-07, "loss": 0.0263, "step": 153350 }, { "epoch": 3.064625714399904, "grad_norm": 0.1529475599527359, "learning_rate": 4.7025951474800046e-07, "loss": 0.032, "step": 153360 }, { "epoch": 3.0648255465409058, "grad_norm": 0.10285313427448273, "learning_rate": 4.688465658083008e-07, "loss": 0.0196, "step": 153370 }, { "epoch": 3.065025378681907, "grad_norm": 0.10599172860383987, "learning_rate": 4.6743573274848907e-07, "loss": 0.0241, "step": 153380 }, { "epoch": 3.0652252108229088, "grad_norm": 0.2187364399433136, "learning_rate": 4.6602701562882823e-07, "loss": 0.021, "step": 153390 }, { "epoch": 3.0654250429639105, "grad_norm": 0.10234753042459488, "learning_rate": 4.646204145094979e-07, "loss": 0.0281, "step": 153400 }, { "epoch": 3.0656248751049118, "grad_norm": 0.13561172783374786, "learning_rate": 4.6321592945058335e-07, "loss": 0.0219, "step": 153410 }, { "epoch": 3.0658247072459135, "grad_norm": 0.15340259671211243, "learning_rate": 4.6181356051208105e-07, "loss": 0.0181, "step": 153420 }, { "epoch": 3.066024539386915, "grad_norm": 0.16782045364379883, "learning_rate": 4.6041330775389856e-07, "loss": 0.021, "step": 153430 }, { "epoch": 3.0662243715279165, "grad_norm": 0.07349663227796555, "learning_rate": 4.590151712358548e-07, "loss": 0.0225, "step": 153440 }, { "epoch": 3.066424203668918, "grad_norm": 0.10888867825269699, "learning_rate": 4.576191510176686e-07, "loss": 0.0211, "step": 153450 }, { "epoch": 3.0666240358099195, "grad_norm": 0.13808311522006989, "learning_rate": 4.5622524715898673e-07, "loss": 0.0193, "step": 153460 }, { "epoch": 3.066823867950921, "grad_norm": 0.1328871101140976, "learning_rate": 4.5483345971934485e-07, "loss": 0.0178, "step": 153470 }, { "epoch": 3.067023700091923, "grad_norm": 0.1242695152759552, "learning_rate": 4.534437887581955e-07, "loss": 0.0261, "step": 153480 }, { "epoch": 3.067223532232924, "grad_norm": 0.12198502570390701, "learning_rate": 4.5205623433491326e-07, "loss": 0.0197, "step": 153490 }, { "epoch": 3.067423364373926, "grad_norm": 0.11478420346975327, "learning_rate": 4.506707965087675e-07, "loss": 0.0203, "step": 153500 }, { "epoch": 3.0676231965149277, "grad_norm": 0.10635072737932205, "learning_rate": 4.492874753389331e-07, "loss": 0.0181, "step": 153510 }, { "epoch": 3.067823028655929, "grad_norm": 0.1543886363506317, "learning_rate": 4.4790627088451274e-07, "loss": 0.0196, "step": 153520 }, { "epoch": 3.0680228607969307, "grad_norm": 0.10640306025743484, "learning_rate": 4.4652718320450924e-07, "loss": 0.0257, "step": 153530 }, { "epoch": 3.068222692937932, "grad_norm": 0.3015681803226471, "learning_rate": 4.4515021235783104e-07, "loss": 0.0255, "step": 153540 }, { "epoch": 3.0684225250789336, "grad_norm": 0.10889724642038345, "learning_rate": 4.4377535840329776e-07, "loss": 0.0208, "step": 153550 }, { "epoch": 3.0686223572199354, "grad_norm": 0.0819951519370079, "learning_rate": 4.424026213996513e-07, "loss": 0.0348, "step": 153560 }, { "epoch": 3.0688221893609366, "grad_norm": 0.09589476883411407, "learning_rate": 4.4103200140551695e-07, "loss": 0.02, "step": 153570 }, { "epoch": 3.0690220215019384, "grad_norm": 0.09371127188205719, "learning_rate": 4.396634984794534e-07, "loss": 0.0252, "step": 153580 }, { "epoch": 3.06922185364294, "grad_norm": 0.11921636760234833, "learning_rate": 4.382971126799196e-07, "loss": 0.0191, "step": 153590 }, { "epoch": 3.0694216857839414, "grad_norm": 0.15694977343082428, "learning_rate": 4.369328440652909e-07, "loss": 0.0261, "step": 153600 }, { "epoch": 3.069621517924943, "grad_norm": 0.09884076565504074, "learning_rate": 4.355706926938319e-07, "loss": 0.0181, "step": 153610 }, { "epoch": 3.069821350065945, "grad_norm": 0.11666245013475418, "learning_rate": 4.34210658623746e-07, "loss": 0.0232, "step": 153620 }, { "epoch": 3.070021182206946, "grad_norm": 0.1106896847486496, "learning_rate": 4.3285274191312566e-07, "loss": 0.0193, "step": 153630 }, { "epoch": 3.070221014347948, "grad_norm": 0.12445122003555298, "learning_rate": 4.3149694261998e-07, "loss": 0.0219, "step": 153640 }, { "epoch": 3.070420846488949, "grad_norm": 0.17868997156620026, "learning_rate": 4.3014326080222377e-07, "loss": 0.0229, "step": 153650 }, { "epoch": 3.070620678629951, "grad_norm": 0.10765527188777924, "learning_rate": 4.28791696517683e-07, "loss": 0.0234, "step": 153660 }, { "epoch": 3.0708205107709525, "grad_norm": 0.08587438613176346, "learning_rate": 4.2744224982409485e-07, "loss": 0.0224, "step": 153670 }, { "epoch": 3.071020342911954, "grad_norm": 0.15699432790279388, "learning_rate": 4.260949207791076e-07, "loss": 0.0285, "step": 153680 }, { "epoch": 3.0712201750529555, "grad_norm": 0.11493799835443497, "learning_rate": 4.247497094402753e-07, "loss": 0.0192, "step": 153690 }, { "epoch": 3.0714200071939572, "grad_norm": 0.1296987235546112, "learning_rate": 4.234066158650574e-07, "loss": 0.0286, "step": 153700 }, { "epoch": 3.0716198393349585, "grad_norm": 0.09107114374637604, "learning_rate": 4.220656401108414e-07, "loss": 0.0267, "step": 153710 }, { "epoch": 3.0718196714759602, "grad_norm": 0.12810030579566956, "learning_rate": 4.2072678223489813e-07, "loss": 0.0302, "step": 153720 }, { "epoch": 3.072019503616962, "grad_norm": 0.11810111999511719, "learning_rate": 4.1939004229443193e-07, "loss": 0.0276, "step": 153730 }, { "epoch": 3.0722193357579632, "grad_norm": 0.13173136115074158, "learning_rate": 4.1805542034654144e-07, "loss": 0.0189, "step": 153740 }, { "epoch": 3.072419167898965, "grad_norm": 0.10430300980806351, "learning_rate": 4.1672291644823113e-07, "loss": 0.0212, "step": 153750 }, { "epoch": 3.0726190000399662, "grad_norm": 0.14687232673168182, "learning_rate": 4.153925306564388e-07, "loss": 0.0192, "step": 153760 }, { "epoch": 3.072818832180968, "grad_norm": 0.11159051954746246, "learning_rate": 4.140642630279801e-07, "loss": 0.0243, "step": 153770 }, { "epoch": 3.0730186643219697, "grad_norm": 0.10182057321071625, "learning_rate": 4.127381136196096e-07, "loss": 0.0322, "step": 153780 }, { "epoch": 3.073218496462971, "grad_norm": 0.08656264841556549, "learning_rate": 4.114140824879653e-07, "loss": 0.0239, "step": 153790 }, { "epoch": 3.0734183286039727, "grad_norm": 0.10246454179286957, "learning_rate": 4.100921696896187e-07, "loss": 0.0272, "step": 153800 }, { "epoch": 3.0736181607449744, "grad_norm": 0.12006369978189468, "learning_rate": 4.0877237528103575e-07, "loss": 0.0222, "step": 153810 }, { "epoch": 3.0738179928859757, "grad_norm": 0.08379777520895004, "learning_rate": 4.0745469931859345e-07, "loss": 0.0264, "step": 153820 }, { "epoch": 3.0740178250269774, "grad_norm": 0.11469414085149765, "learning_rate": 4.061391418585858e-07, "loss": 0.0233, "step": 153830 }, { "epoch": 3.0742176571679787, "grad_norm": 0.1357385516166687, "learning_rate": 4.048257029572011e-07, "loss": 0.0264, "step": 153840 }, { "epoch": 3.0744174893089804, "grad_norm": 0.1223536878824234, "learning_rate": 4.0351438267055565e-07, "loss": 0.0182, "step": 153850 }, { "epoch": 3.074617321449982, "grad_norm": 0.13564680516719818, "learning_rate": 4.0220518105466564e-07, "loss": 0.0223, "step": 153860 }, { "epoch": 3.0748171535909834, "grad_norm": 0.11415320634841919, "learning_rate": 4.0089809816545865e-07, "loss": 0.0232, "step": 153870 }, { "epoch": 3.075016985731985, "grad_norm": 0.10149437934160233, "learning_rate": 3.995931340587622e-07, "loss": 0.0261, "step": 153880 }, { "epoch": 3.075216817872987, "grad_norm": 0.10915812104940414, "learning_rate": 3.982902887903317e-07, "loss": 0.0243, "step": 153890 }, { "epoch": 3.075416650013988, "grad_norm": 0.09243310242891312, "learning_rate": 3.9698956241581706e-07, "loss": 0.0224, "step": 153900 }, { "epoch": 3.07561648215499, "grad_norm": 0.1399718075990677, "learning_rate": 3.95690954990785e-07, "loss": 0.023, "step": 153910 }, { "epoch": 3.0758163142959916, "grad_norm": 0.1516374796628952, "learning_rate": 3.943944665707133e-07, "loss": 0.0281, "step": 153920 }, { "epoch": 3.076016146436993, "grad_norm": 0.13670453429222107, "learning_rate": 3.931000972109744e-07, "loss": 0.0271, "step": 153930 }, { "epoch": 3.0762159785779946, "grad_norm": 0.13289889693260193, "learning_rate": 3.91807846966874e-07, "loss": 0.0232, "step": 153940 }, { "epoch": 3.076415810718996, "grad_norm": 0.10949164628982544, "learning_rate": 3.9051771589360684e-07, "loss": 0.0284, "step": 153950 }, { "epoch": 3.0766156428599976, "grad_norm": 0.13170084357261658, "learning_rate": 3.8922970404629e-07, "loss": 0.0179, "step": 153960 }, { "epoch": 3.0768154750009993, "grad_norm": 0.14168472588062286, "learning_rate": 3.8794381147994054e-07, "loss": 0.0256, "step": 153970 }, { "epoch": 3.0770153071420006, "grad_norm": 0.11613380908966064, "learning_rate": 3.866600382494867e-07, "loss": 0.0225, "step": 153980 }, { "epoch": 3.0772151392830023, "grad_norm": 0.08466251194477081, "learning_rate": 3.853783844097791e-07, "loss": 0.0167, "step": 153990 }, { "epoch": 3.077414971424004, "grad_norm": 0.07094576954841614, "learning_rate": 3.8409885001556287e-07, "loss": 0.0253, "step": 154000 }, { "epoch": 3.0776148035650053, "grad_norm": 0.09466291218996048, "learning_rate": 3.828214351214943e-07, "loss": 0.0208, "step": 154010 }, { "epoch": 3.077814635706007, "grad_norm": 0.17974182963371277, "learning_rate": 3.815461397821407e-07, "loss": 0.0246, "step": 154020 }, { "epoch": 3.0780144678470087, "grad_norm": 0.12034304440021515, "learning_rate": 3.8027296405199195e-07, "loss": 0.0227, "step": 154030 }, { "epoch": 3.07821429998801, "grad_norm": 0.1415303498506546, "learning_rate": 3.790019079854212e-07, "loss": 0.0288, "step": 154040 }, { "epoch": 3.0784141321290117, "grad_norm": 0.1276993453502655, "learning_rate": 3.7773297163673506e-07, "loss": 0.0245, "step": 154050 }, { "epoch": 3.078613964270013, "grad_norm": 0.10432948917150497, "learning_rate": 3.764661550601345e-07, "loss": 0.027, "step": 154060 }, { "epoch": 3.0788137964110147, "grad_norm": 0.15007592737674713, "learning_rate": 3.75201458309743e-07, "loss": 0.0203, "step": 154070 }, { "epoch": 3.0790136285520164, "grad_norm": 0.12036403268575668, "learning_rate": 3.739388814395783e-07, "loss": 0.0257, "step": 154080 }, { "epoch": 3.0792134606930177, "grad_norm": 0.10885074734687805, "learning_rate": 3.726784245035808e-07, "loss": 0.0226, "step": 154090 }, { "epoch": 3.0794132928340194, "grad_norm": 0.12916047871112823, "learning_rate": 3.7142008755559064e-07, "loss": 0.0305, "step": 154100 }, { "epoch": 3.079613124975021, "grad_norm": 0.125139057636261, "learning_rate": 3.701638706493649e-07, "loss": 0.0237, "step": 154110 }, { "epoch": 3.0798129571160224, "grad_norm": 0.11478138715028763, "learning_rate": 3.6890977383856607e-07, "loss": 0.0231, "step": 154120 }, { "epoch": 3.080012789257024, "grad_norm": 0.08420324325561523, "learning_rate": 3.6765779717676805e-07, "loss": 0.024, "step": 154130 }, { "epoch": 3.080212621398026, "grad_norm": 0.11529319733381271, "learning_rate": 3.664079407174448e-07, "loss": 0.0247, "step": 154140 }, { "epoch": 3.080412453539027, "grad_norm": 0.10372646898031235, "learning_rate": 3.6516020451399793e-07, "loss": 0.019, "step": 154150 }, { "epoch": 3.080612285680029, "grad_norm": 0.13980785012245178, "learning_rate": 3.639145886197293e-07, "loss": 0.0217, "step": 154160 }, { "epoch": 3.08081211782103, "grad_norm": 0.08266110718250275, "learning_rate": 3.6267109308784076e-07, "loss": 0.0264, "step": 154170 }, { "epoch": 3.081011949962032, "grad_norm": 0.1168157309293747, "learning_rate": 3.6142971797145654e-07, "loss": 0.0237, "step": 154180 }, { "epoch": 3.0812117821030336, "grad_norm": 0.08461114019155502, "learning_rate": 3.601904633236064e-07, "loss": 0.0234, "step": 154190 }, { "epoch": 3.081411614244035, "grad_norm": 0.15300379693508148, "learning_rate": 3.5895332919723133e-07, "loss": 0.0279, "step": 154200 }, { "epoch": 3.0816114463850366, "grad_norm": 0.11972937732934952, "learning_rate": 3.5771831564517243e-07, "loss": 0.0262, "step": 154210 }, { "epoch": 3.0818112785260383, "grad_norm": 0.1933426707983017, "learning_rate": 3.5648542272019305e-07, "loss": 0.0247, "step": 154220 }, { "epoch": 3.0820111106670396, "grad_norm": 0.10731958597898483, "learning_rate": 3.5525465047495656e-07, "loss": 0.024, "step": 154230 }, { "epoch": 3.0822109428080413, "grad_norm": 0.1297818273305893, "learning_rate": 3.540259989620431e-07, "loss": 0.0267, "step": 154240 }, { "epoch": 3.082410774949043, "grad_norm": 0.1041308045387268, "learning_rate": 3.5279946823393304e-07, "loss": 0.0238, "step": 154250 }, { "epoch": 3.0826106070900443, "grad_norm": 0.1457180678844452, "learning_rate": 3.5157505834302886e-07, "loss": 0.0194, "step": 154260 }, { "epoch": 3.082810439231046, "grad_norm": 0.11468765884637833, "learning_rate": 3.5035276934163305e-07, "loss": 0.0258, "step": 154270 }, { "epoch": 3.0830102713720473, "grad_norm": 0.7795820236206055, "learning_rate": 3.491326012819485e-07, "loss": 0.0237, "step": 154280 }, { "epoch": 3.083210103513049, "grad_norm": 0.092849500477314, "learning_rate": 3.4791455421611664e-07, "loss": 0.0238, "step": 154290 }, { "epoch": 3.0834099356540507, "grad_norm": 0.10295087844133377, "learning_rate": 3.4669862819615705e-07, "loss": 0.0174, "step": 154300 }, { "epoch": 3.083609767795052, "grad_norm": 0.09793485701084137, "learning_rate": 3.454848232740171e-07, "loss": 0.0208, "step": 154310 }, { "epoch": 3.0838095999360537, "grad_norm": 0.10197211056947708, "learning_rate": 3.442731395015497e-07, "loss": 0.0247, "step": 154320 }, { "epoch": 3.0840094320770555, "grad_norm": 0.13227073848247528, "learning_rate": 3.4306357693050794e-07, "loss": 0.027, "step": 154330 }, { "epoch": 3.0842092642180567, "grad_norm": 0.11860258877277374, "learning_rate": 3.4185613561257267e-07, "loss": 0.0296, "step": 154340 }, { "epoch": 3.0844090963590585, "grad_norm": 0.11529558897018433, "learning_rate": 3.406508155993138e-07, "loss": 0.0243, "step": 154350 }, { "epoch": 3.0846089285000597, "grad_norm": 0.10580497235059738, "learning_rate": 3.3944761694222894e-07, "loss": 0.0255, "step": 154360 }, { "epoch": 3.0848087606410615, "grad_norm": 0.13175621628761292, "learning_rate": 3.3824653969271037e-07, "loss": 0.0212, "step": 154370 }, { "epoch": 3.085008592782063, "grad_norm": 0.12375613301992416, "learning_rate": 3.3704758390206704e-07, "loss": 0.0225, "step": 154380 }, { "epoch": 3.0852084249230645, "grad_norm": 0.10188992321491241, "learning_rate": 3.358507496215191e-07, "loss": 0.0314, "step": 154390 }, { "epoch": 3.085408257064066, "grad_norm": 0.13375356793403625, "learning_rate": 3.346560369021923e-07, "loss": 0.0246, "step": 154400 }, { "epoch": 3.085608089205068, "grad_norm": 0.08330906182527542, "learning_rate": 3.33463445795118e-07, "loss": 0.0255, "step": 154410 }, { "epoch": 3.085807921346069, "grad_norm": 0.13563603162765503, "learning_rate": 3.3227297635124444e-07, "loss": 0.0246, "step": 154420 }, { "epoch": 3.086007753487071, "grad_norm": 0.11816062778234482, "learning_rate": 3.3108462862143083e-07, "loss": 0.0267, "step": 154430 }, { "epoch": 3.0862075856280726, "grad_norm": 0.10525273531675339, "learning_rate": 3.298984026564367e-07, "loss": 0.0303, "step": 154440 }, { "epoch": 3.086407417769074, "grad_norm": 0.11612270772457123, "learning_rate": 3.28714298506938e-07, "loss": 0.0218, "step": 154450 }, { "epoch": 3.0866072499100756, "grad_norm": 0.09726964682340622, "learning_rate": 3.275323162235111e-07, "loss": 0.0198, "step": 154460 }, { "epoch": 3.086807082051077, "grad_norm": 0.16383294761180878, "learning_rate": 3.263524558566544e-07, "loss": 0.0204, "step": 154470 }, { "epoch": 3.0870069141920786, "grad_norm": 0.10145706683397293, "learning_rate": 3.2517471745676654e-07, "loss": 0.0251, "step": 154480 }, { "epoch": 3.0872067463330803, "grad_norm": 0.12193029373884201, "learning_rate": 3.239991010741572e-07, "loss": 0.0269, "step": 154490 }, { "epoch": 3.0874065784740816, "grad_norm": 0.15447135269641876, "learning_rate": 3.2282560675905294e-07, "loss": 0.0237, "step": 154500 }, { "epoch": 3.0876064106150833, "grad_norm": 0.11516183614730835, "learning_rate": 3.216542345615747e-07, "loss": 0.0183, "step": 154510 }, { "epoch": 3.087806242756085, "grad_norm": 0.1512092798948288, "learning_rate": 3.204849845317659e-07, "loss": 0.0141, "step": 154520 }, { "epoch": 3.0880060748970863, "grad_norm": 0.13415955007076263, "learning_rate": 3.193178567195754e-07, "loss": 0.0225, "step": 154530 }, { "epoch": 3.088205907038088, "grad_norm": 0.11553876101970673, "learning_rate": 3.181528511748633e-07, "loss": 0.0244, "step": 154540 }, { "epoch": 3.0884057391790893, "grad_norm": 0.13165901601314545, "learning_rate": 3.1698996794738424e-07, "loss": 0.0206, "step": 154550 }, { "epoch": 3.088605571320091, "grad_norm": 0.13664035499095917, "learning_rate": 3.1582920708683186e-07, "loss": 0.0274, "step": 154560 }, { "epoch": 3.088805403461093, "grad_norm": 0.1037345677614212, "learning_rate": 3.1467056864277756e-07, "loss": 0.0244, "step": 154570 }, { "epoch": 3.089005235602094, "grad_norm": 0.11854493618011475, "learning_rate": 3.1351405266472625e-07, "loss": 0.0217, "step": 154580 }, { "epoch": 3.089205067743096, "grad_norm": 0.16324001550674438, "learning_rate": 3.123596592020717e-07, "loss": 0.0286, "step": 154590 }, { "epoch": 3.0894048998840975, "grad_norm": 0.18935011327266693, "learning_rate": 3.112073883041411e-07, "loss": 0.0272, "step": 154600 }, { "epoch": 3.0896047320250988, "grad_norm": 0.07472965866327286, "learning_rate": 3.1005724002014515e-07, "loss": 0.025, "step": 154610 }, { "epoch": 3.0898045641661005, "grad_norm": 0.13047559559345245, "learning_rate": 3.089092143992167e-07, "loss": 0.0213, "step": 154620 }, { "epoch": 3.090004396307102, "grad_norm": 0.1289927214384079, "learning_rate": 3.0776331149041105e-07, "loss": 0.0206, "step": 154630 }, { "epoch": 3.0902042284481035, "grad_norm": 0.13890191912651062, "learning_rate": 3.066195313426612e-07, "loss": 0.0239, "step": 154640 }, { "epoch": 3.090404060589105, "grad_norm": 0.09338589757680893, "learning_rate": 3.054778740048336e-07, "loss": 0.0206, "step": 154650 }, { "epoch": 3.0906038927301065, "grad_norm": 0.09924937784671783, "learning_rate": 3.04338339525706e-07, "loss": 0.0204, "step": 154660 }, { "epoch": 3.090803724871108, "grad_norm": 0.12041613459587097, "learning_rate": 3.032009279539505e-07, "loss": 0.0238, "step": 154670 }, { "epoch": 3.09100355701211, "grad_norm": 0.0945177897810936, "learning_rate": 3.0206563933815044e-07, "loss": 0.0239, "step": 154680 }, { "epoch": 3.091203389153111, "grad_norm": 0.0903443992137909, "learning_rate": 3.0093247372681156e-07, "loss": 0.0304, "step": 154690 }, { "epoch": 3.091403221294113, "grad_norm": 0.1041409820318222, "learning_rate": 2.998014311683339e-07, "loss": 0.0247, "step": 154700 }, { "epoch": 3.0916030534351147, "grad_norm": 0.19621634483337402, "learning_rate": 2.9867251171104e-07, "loss": 0.0301, "step": 154710 }, { "epoch": 3.091802885576116, "grad_norm": 0.13356110453605652, "learning_rate": 2.9754571540315246e-07, "loss": 0.0244, "step": 154720 }, { "epoch": 3.0920027177171177, "grad_norm": 0.11588185280561447, "learning_rate": 2.9642104229280486e-07, "loss": 0.0144, "step": 154730 }, { "epoch": 3.0922025498581194, "grad_norm": 0.07660423219203949, "learning_rate": 2.952984924280422e-07, "loss": 0.0217, "step": 154740 }, { "epoch": 3.0924023819991207, "grad_norm": 0.10952131450176239, "learning_rate": 2.9417806585681493e-07, "loss": 0.0226, "step": 154750 }, { "epoch": 3.0926022141401224, "grad_norm": 0.17454956471920013, "learning_rate": 2.930597626269904e-07, "loss": 0.0249, "step": 154760 }, { "epoch": 3.0928020462811237, "grad_norm": 0.12194360047578812, "learning_rate": 2.919435827863415e-07, "loss": 0.0211, "step": 154770 }, { "epoch": 3.0930018784221254, "grad_norm": 0.12871405482292175, "learning_rate": 2.9082952638254114e-07, "loss": 0.0229, "step": 154780 }, { "epoch": 3.093201710563127, "grad_norm": 0.14206674695014954, "learning_rate": 2.897175934631846e-07, "loss": 0.0314, "step": 154790 }, { "epoch": 3.0934015427041284, "grad_norm": 0.13170552253723145, "learning_rate": 2.8860778407577837e-07, "loss": 0.025, "step": 154800 }, { "epoch": 3.09360137484513, "grad_norm": 0.19842800498008728, "learning_rate": 2.875000982677178e-07, "loss": 0.0207, "step": 154810 }, { "epoch": 3.093801206986132, "grad_norm": 0.12186568230390549, "learning_rate": 2.863945360863318e-07, "loss": 0.0232, "step": 154820 }, { "epoch": 3.094001039127133, "grad_norm": 0.07907138019800186, "learning_rate": 2.8529109757884367e-07, "loss": 0.0165, "step": 154830 }, { "epoch": 3.094200871268135, "grad_norm": 0.14955958724021912, "learning_rate": 2.8418978279238784e-07, "loss": 0.0225, "step": 154840 }, { "epoch": 3.0944007034091365, "grad_norm": 0.09870343655347824, "learning_rate": 2.8309059177401564e-07, "loss": 0.0241, "step": 154850 }, { "epoch": 3.094600535550138, "grad_norm": 0.11272697895765305, "learning_rate": 2.8199352457068397e-07, "loss": 0.0218, "step": 154860 }, { "epoch": 3.0948003676911395, "grad_norm": 0.10214733332395554, "learning_rate": 2.808985812292497e-07, "loss": 0.0202, "step": 154870 }, { "epoch": 3.095000199832141, "grad_norm": 0.16110754013061523, "learning_rate": 2.7980576179649223e-07, "loss": 0.0242, "step": 154880 }, { "epoch": 3.0952000319731425, "grad_norm": 0.12694057822227478, "learning_rate": 2.7871506631909626e-07, "loss": 0.03, "step": 154890 }, { "epoch": 3.0953998641141443, "grad_norm": 0.11286401003599167, "learning_rate": 2.7762649484365246e-07, "loss": 0.0206, "step": 154900 }, { "epoch": 3.0955996962551455, "grad_norm": 0.13992822170257568, "learning_rate": 2.7654004741665686e-07, "loss": 0.0252, "step": 154910 }, { "epoch": 3.0957995283961472, "grad_norm": 0.09245345741510391, "learning_rate": 2.754557240845279e-07, "loss": 0.0261, "step": 154920 }, { "epoch": 3.095999360537149, "grad_norm": 0.11777833104133606, "learning_rate": 2.7437352489358414e-07, "loss": 0.0214, "step": 154930 }, { "epoch": 3.0961991926781502, "grad_norm": 0.11501853913068771, "learning_rate": 2.7329344989005523e-07, "loss": 0.0312, "step": 154940 }, { "epoch": 3.096399024819152, "grad_norm": 0.13237585127353668, "learning_rate": 2.7221549912007647e-07, "loss": 0.0237, "step": 154950 }, { "epoch": 3.0965988569601537, "grad_norm": 0.10292977094650269, "learning_rate": 2.7113967262969997e-07, "loss": 0.0247, "step": 154960 }, { "epoch": 3.096798689101155, "grad_norm": 0.1018889993429184, "learning_rate": 2.700659704648778e-07, "loss": 0.0256, "step": 154970 }, { "epoch": 3.0969985212421567, "grad_norm": 0.09279246628284454, "learning_rate": 2.689943926714844e-07, "loss": 0.0277, "step": 154980 }, { "epoch": 3.097198353383158, "grad_norm": 0.11157698929309845, "learning_rate": 2.6792493929529427e-07, "loss": 0.0213, "step": 154990 }, { "epoch": 3.0973981855241597, "grad_norm": 0.14264117181301117, "learning_rate": 2.668576103819875e-07, "loss": 0.0306, "step": 155000 }, { "epoch": 3.0975980176651614, "grad_norm": 0.11565393209457397, "learning_rate": 2.65792405977161e-07, "loss": 0.0229, "step": 155010 }, { "epoch": 3.0977978498061627, "grad_norm": 0.13176897168159485, "learning_rate": 2.647293261263173e-07, "loss": 0.0231, "step": 155020 }, { "epoch": 3.0979976819471644, "grad_norm": 0.09962207078933716, "learning_rate": 2.636683708748755e-07, "loss": 0.0281, "step": 155030 }, { "epoch": 3.098197514088166, "grad_norm": 0.10705117881298065, "learning_rate": 2.6260954026814386e-07, "loss": 0.0197, "step": 155040 }, { "epoch": 3.0983973462291674, "grad_norm": 0.09103766083717346, "learning_rate": 2.61552834351364e-07, "loss": 0.0189, "step": 155050 }, { "epoch": 3.098597178370169, "grad_norm": 0.13310882449150085, "learning_rate": 2.6049825316968314e-07, "loss": 0.0318, "step": 155060 }, { "epoch": 3.0987970105111704, "grad_norm": 0.11044290661811829, "learning_rate": 2.594457967681374e-07, "loss": 0.0184, "step": 155070 }, { "epoch": 3.098996842652172, "grad_norm": 0.11306481063365936, "learning_rate": 2.5839546519168535e-07, "loss": 0.018, "step": 155080 }, { "epoch": 3.099196674793174, "grad_norm": 0.11985012888908386, "learning_rate": 2.573472584852077e-07, "loss": 0.0193, "step": 155090 }, { "epoch": 3.099396506934175, "grad_norm": 0.14490053057670593, "learning_rate": 2.563011766934742e-07, "loss": 0.0234, "step": 155100 }, { "epoch": 3.099596339075177, "grad_norm": 0.12422055751085281, "learning_rate": 2.552572198611658e-07, "loss": 0.0274, "step": 155110 }, { "epoch": 3.0997961712161786, "grad_norm": 0.10982096940279007, "learning_rate": 2.5421538803289125e-07, "loss": 0.0181, "step": 155120 }, { "epoch": 3.09999600335718, "grad_norm": 0.14112339913845062, "learning_rate": 2.531756812531483e-07, "loss": 0.0229, "step": 155130 }, { "epoch": 3.1001958354981816, "grad_norm": 0.13645435869693756, "learning_rate": 2.521380995663514e-07, "loss": 0.0249, "step": 155140 }, { "epoch": 3.1003956676391833, "grad_norm": 0.09805995225906372, "learning_rate": 2.511026430168262e-07, "loss": 0.0251, "step": 155150 }, { "epoch": 3.1005954997801846, "grad_norm": 0.10385701060295105, "learning_rate": 2.50069311648804e-07, "loss": 0.0174, "step": 155160 }, { "epoch": 3.1007953319211863, "grad_norm": 0.12137501686811447, "learning_rate": 2.4903810550642725e-07, "loss": 0.0165, "step": 155170 }, { "epoch": 3.1009951640621876, "grad_norm": 0.15941767394542694, "learning_rate": 2.4800902463374964e-07, "loss": 0.022, "step": 155180 }, { "epoch": 3.1011949962031893, "grad_norm": 0.13346315920352936, "learning_rate": 2.4698206907472487e-07, "loss": 0.0213, "step": 155190 }, { "epoch": 3.101394828344191, "grad_norm": 0.16610406339168549, "learning_rate": 2.45957238873229e-07, "loss": 0.0256, "step": 155200 }, { "epoch": 3.1015946604851923, "grad_norm": 0.1314600706100464, "learning_rate": 2.44934534073038e-07, "loss": 0.0241, "step": 155210 }, { "epoch": 3.101794492626194, "grad_norm": 0.10707981884479523, "learning_rate": 2.4391395471783926e-07, "loss": 0.0248, "step": 155220 }, { "epoch": 3.1019943247671957, "grad_norm": 0.13228845596313477, "learning_rate": 2.428955008512368e-07, "loss": 0.0248, "step": 155230 }, { "epoch": 3.102194156908197, "grad_norm": 0.09706642478704453, "learning_rate": 2.4187917251672907e-07, "loss": 0.0197, "step": 155240 }, { "epoch": 3.1023939890491987, "grad_norm": 0.13005436956882477, "learning_rate": 2.4086496975773145e-07, "loss": 0.0302, "step": 155250 }, { "epoch": 3.1025938211902004, "grad_norm": 0.12002814561128616, "learning_rate": 2.3985289261757583e-07, "loss": 0.0217, "step": 155260 }, { "epoch": 3.1027936533312017, "grad_norm": 0.11504649370908737, "learning_rate": 2.388429411394943e-07, "loss": 0.0226, "step": 155270 }, { "epoch": 3.1029934854722034, "grad_norm": 0.07865295559167862, "learning_rate": 2.3783511536662472e-07, "loss": 0.0199, "step": 155280 }, { "epoch": 3.1031933176132047, "grad_norm": 0.10009744018316269, "learning_rate": 2.368294153420214e-07, "loss": 0.0325, "step": 155290 }, { "epoch": 3.1033931497542064, "grad_norm": 0.15380410850048065, "learning_rate": 2.3582584110865002e-07, "loss": 0.0275, "step": 155300 }, { "epoch": 3.103592981895208, "grad_norm": 0.1192798912525177, "learning_rate": 2.3482439270937628e-07, "loss": 0.0247, "step": 155310 }, { "epoch": 3.1037928140362094, "grad_norm": 0.10051897168159485, "learning_rate": 2.3382507018698264e-07, "loss": 0.0199, "step": 155320 }, { "epoch": 3.103992646177211, "grad_norm": 0.10617726296186447, "learning_rate": 2.328278735841627e-07, "loss": 0.0225, "step": 155330 }, { "epoch": 3.104192478318213, "grad_norm": 0.07336621731519699, "learning_rate": 2.3183280294351016e-07, "loss": 0.021, "step": 155340 }, { "epoch": 3.104392310459214, "grad_norm": 0.11488448828458786, "learning_rate": 2.3083985830753552e-07, "loss": 0.0284, "step": 155350 }, { "epoch": 3.104592142600216, "grad_norm": 0.10626048594713211, "learning_rate": 2.2984903971864924e-07, "loss": 0.0205, "step": 155360 }, { "epoch": 3.104791974741217, "grad_norm": 0.10522043704986572, "learning_rate": 2.288603472191786e-07, "loss": 0.0161, "step": 155370 }, { "epoch": 3.104991806882219, "grad_norm": 0.09702872484922409, "learning_rate": 2.2787378085136757e-07, "loss": 0.0206, "step": 155380 }, { "epoch": 3.1051916390232206, "grad_norm": 0.11906765401363373, "learning_rate": 2.2688934065734913e-07, "loss": 0.0356, "step": 155390 }, { "epoch": 3.105391471164222, "grad_norm": 0.10247678309679031, "learning_rate": 2.2590702667918405e-07, "loss": 0.0193, "step": 155400 }, { "epoch": 3.1055913033052236, "grad_norm": 0.15657155215740204, "learning_rate": 2.2492683895883326e-07, "loss": 0.0191, "step": 155410 }, { "epoch": 3.1057911354462253, "grad_norm": 0.09972470998764038, "learning_rate": 2.239487775381688e-07, "loss": 0.0192, "step": 155420 }, { "epoch": 3.1059909675872266, "grad_norm": 0.11791497468948364, "learning_rate": 2.2297284245896833e-07, "loss": 0.0232, "step": 155430 }, { "epoch": 3.1061907997282283, "grad_norm": 0.09843869507312775, "learning_rate": 2.2199903376293186e-07, "loss": 0.0164, "step": 155440 }, { "epoch": 3.10639063186923, "grad_norm": 0.10802841931581497, "learning_rate": 2.2102735149164277e-07, "loss": 0.0257, "step": 155450 }, { "epoch": 3.1065904640102313, "grad_norm": 0.12153242528438568, "learning_rate": 2.2005779568662343e-07, "loss": 0.0298, "step": 155460 }, { "epoch": 3.106790296151233, "grad_norm": 0.1177699938416481, "learning_rate": 2.1909036638928516e-07, "loss": 0.0208, "step": 155470 }, { "epoch": 3.1069901282922343, "grad_norm": 0.11481406539678574, "learning_rate": 2.1812506364095597e-07, "loss": 0.0222, "step": 155480 }, { "epoch": 3.107189960433236, "grad_norm": 0.12001527100801468, "learning_rate": 2.171618874828696e-07, "loss": 0.0271, "step": 155490 }, { "epoch": 3.1073897925742378, "grad_norm": 0.11223828792572021, "learning_rate": 2.162008379561764e-07, "loss": 0.02, "step": 155500 }, { "epoch": 3.107589624715239, "grad_norm": 0.1503266990184784, "learning_rate": 2.1524191510192694e-07, "loss": 0.0297, "step": 155510 }, { "epoch": 3.1077894568562408, "grad_norm": 0.10565263032913208, "learning_rate": 2.1428511896108283e-07, "loss": 0.0231, "step": 155520 }, { "epoch": 3.1079892889972425, "grad_norm": 0.08985686302185059, "learning_rate": 2.1333044957452252e-07, "loss": 0.0159, "step": 155530 }, { "epoch": 3.1081891211382437, "grad_norm": 0.09081030637025833, "learning_rate": 2.1237790698301895e-07, "loss": 0.0245, "step": 155540 }, { "epoch": 3.1083889532792455, "grad_norm": 0.10671256482601166, "learning_rate": 2.114274912272729e-07, "loss": 0.0289, "step": 155550 }, { "epoch": 3.108588785420247, "grad_norm": 0.09828364104032516, "learning_rate": 2.1047920234787967e-07, "loss": 0.0185, "step": 155560 }, { "epoch": 3.1087886175612485, "grad_norm": 0.08512412011623383, "learning_rate": 2.0953304038535127e-07, "loss": 0.0278, "step": 155570 }, { "epoch": 3.10898844970225, "grad_norm": 0.10426271706819534, "learning_rate": 2.0858900538009985e-07, "loss": 0.0273, "step": 155580 }, { "epoch": 3.1091882818432515, "grad_norm": 0.11514141410589218, "learning_rate": 2.0764709737245424e-07, "loss": 0.0185, "step": 155590 }, { "epoch": 3.109388113984253, "grad_norm": 0.09624540060758591, "learning_rate": 2.0670731640266007e-07, "loss": 0.019, "step": 155600 }, { "epoch": 3.109587946125255, "grad_norm": 0.10371341556310654, "learning_rate": 2.0576966251085183e-07, "loss": 0.0263, "step": 155610 }, { "epoch": 3.109787778266256, "grad_norm": 0.13781660795211792, "learning_rate": 2.0483413573708644e-07, "loss": 0.0222, "step": 155620 }, { "epoch": 3.109987610407258, "grad_norm": 0.11291369050741196, "learning_rate": 2.0390073612133743e-07, "loss": 0.0244, "step": 155630 }, { "epoch": 3.1101874425482596, "grad_norm": 0.09247252345085144, "learning_rate": 2.0296946370346182e-07, "loss": 0.0252, "step": 155640 }, { "epoch": 3.110387274689261, "grad_norm": 0.09729593247175217, "learning_rate": 2.0204031852325555e-07, "loss": 0.0168, "step": 155650 }, { "epoch": 3.1105871068302626, "grad_norm": 0.0897347554564476, "learning_rate": 2.0111330062040358e-07, "loss": 0.0245, "step": 155660 }, { "epoch": 3.1107869389712643, "grad_norm": 0.16128350794315338, "learning_rate": 2.001884100345075e-07, "loss": 0.025, "step": 155670 }, { "epoch": 3.1109867711122656, "grad_norm": 0.1011873260140419, "learning_rate": 1.992656468050802e-07, "loss": 0.0208, "step": 155680 }, { "epoch": 3.1111866032532673, "grad_norm": 0.10139531642198563, "learning_rate": 1.9834501097153456e-07, "loss": 0.0283, "step": 155690 }, { "epoch": 3.1113864353942686, "grad_norm": 0.09035307168960571, "learning_rate": 1.9742650257320584e-07, "loss": 0.0198, "step": 155700 }, { "epoch": 3.1115862675352703, "grad_norm": 0.13080526888370514, "learning_rate": 1.965101216493237e-07, "loss": 0.0257, "step": 155710 }, { "epoch": 3.111786099676272, "grad_norm": 0.09331417083740234, "learning_rate": 1.9559586823903465e-07, "loss": 0.0274, "step": 155720 }, { "epoch": 3.1119859318172733, "grad_norm": 0.0982849970459938, "learning_rate": 1.9468374238140185e-07, "loss": 0.0217, "step": 155730 }, { "epoch": 3.112185763958275, "grad_norm": 0.1418396532535553, "learning_rate": 1.9377374411538306e-07, "loss": 0.0295, "step": 155740 }, { "epoch": 3.112385596099277, "grad_norm": 0.1278592050075531, "learning_rate": 1.928658734798472e-07, "loss": 0.0274, "step": 155750 }, { "epoch": 3.112585428240278, "grad_norm": 0.17337246239185333, "learning_rate": 1.9196013051359096e-07, "loss": 0.0173, "step": 155760 }, { "epoch": 3.11278526038128, "grad_norm": 0.13408510386943817, "learning_rate": 1.91056515255289e-07, "loss": 0.0254, "step": 155770 }, { "epoch": 3.1129850925222815, "grad_norm": 0.15296582877635956, "learning_rate": 1.9015502774355487e-07, "loss": 0.0206, "step": 155780 }, { "epoch": 3.113184924663283, "grad_norm": 0.10894168168306351, "learning_rate": 1.8925566801689664e-07, "loss": 0.0168, "step": 155790 }, { "epoch": 3.1133847568042845, "grad_norm": 0.13478127121925354, "learning_rate": 1.8835843611372806e-07, "loss": 0.0246, "step": 155800 }, { "epoch": 3.113584588945286, "grad_norm": 0.10183608531951904, "learning_rate": 1.8746333207237955e-07, "loss": 0.0203, "step": 155810 }, { "epoch": 3.1137844210862875, "grad_norm": 0.08773022145032883, "learning_rate": 1.8657035593109273e-07, "loss": 0.0201, "step": 155820 }, { "epoch": 3.1139842532272892, "grad_norm": 0.10779379308223724, "learning_rate": 1.8567950772800935e-07, "loss": 0.0219, "step": 155830 }, { "epoch": 3.1141840853682905, "grad_norm": 0.13978223502635956, "learning_rate": 1.8479078750118227e-07, "loss": 0.0293, "step": 155840 }, { "epoch": 3.114383917509292, "grad_norm": 0.10988636314868927, "learning_rate": 1.8390419528858116e-07, "loss": 0.0182, "step": 155850 }, { "epoch": 3.114583749650294, "grad_norm": 0.0748710036277771, "learning_rate": 1.8301973112808124e-07, "loss": 0.0198, "step": 155860 }, { "epoch": 3.114783581791295, "grad_norm": 0.11081502586603165, "learning_rate": 1.8213739505745786e-07, "loss": 0.0233, "step": 155870 }, { "epoch": 3.114983413932297, "grad_norm": 0.12138956785202026, "learning_rate": 1.8125718711440309e-07, "loss": 0.0192, "step": 155880 }, { "epoch": 3.115183246073298, "grad_norm": 0.13867498934268951, "learning_rate": 1.8037910733652573e-07, "loss": 0.0233, "step": 155890 }, { "epoch": 3.1153830782143, "grad_norm": 0.11262780427932739, "learning_rate": 1.7950315576132914e-07, "loss": 0.0192, "step": 155900 }, { "epoch": 3.1155829103553017, "grad_norm": 0.11180604994297028, "learning_rate": 1.786293324262389e-07, "loss": 0.0233, "step": 155910 }, { "epoch": 3.115782742496303, "grad_norm": 0.11799643933773041, "learning_rate": 1.777576373685752e-07, "loss": 0.0173, "step": 155920 }, { "epoch": 3.1159825746373047, "grad_norm": 0.09660281985998154, "learning_rate": 1.7688807062558045e-07, "loss": 0.0204, "step": 155930 }, { "epoch": 3.1161824067783064, "grad_norm": 0.1639380156993866, "learning_rate": 1.7602063223439712e-07, "loss": 0.0284, "step": 155940 }, { "epoch": 3.1163822389193077, "grad_norm": 0.12627317011356354, "learning_rate": 1.7515532223208453e-07, "loss": 0.0368, "step": 155950 }, { "epoch": 3.1165820710603094, "grad_norm": 0.0929470881819725, "learning_rate": 1.742921406556075e-07, "loss": 0.0267, "step": 155960 }, { "epoch": 3.116781903201311, "grad_norm": 0.1178165003657341, "learning_rate": 1.7343108754183656e-07, "loss": 0.0214, "step": 155970 }, { "epoch": 3.1169817353423124, "grad_norm": 0.19011680781841278, "learning_rate": 1.725721629275534e-07, "loss": 0.021, "step": 155980 }, { "epoch": 3.117181567483314, "grad_norm": 0.12666621804237366, "learning_rate": 1.7171536684945089e-07, "loss": 0.0271, "step": 155990 }, { "epoch": 3.1173813996243154, "grad_norm": 0.1325615495443344, "learning_rate": 1.7086069934413308e-07, "loss": 0.0249, "step": 156000 }, { "epoch": 3.117581231765317, "grad_norm": 0.12327415496110916, "learning_rate": 1.700081604481041e-07, "loss": 0.0236, "step": 156010 }, { "epoch": 3.117781063906319, "grad_norm": 0.08686922490596771, "learning_rate": 1.691577501977848e-07, "loss": 0.0201, "step": 156020 }, { "epoch": 3.11798089604732, "grad_norm": 0.1214241310954094, "learning_rate": 1.6830946862950724e-07, "loss": 0.0184, "step": 156030 }, { "epoch": 3.118180728188322, "grad_norm": 0.1168246790766716, "learning_rate": 1.674633157795036e-07, "loss": 0.0253, "step": 156040 }, { "epoch": 3.1183805603293235, "grad_norm": 0.18053148686885834, "learning_rate": 1.6661929168392266e-07, "loss": 0.0214, "step": 156050 }, { "epoch": 3.118580392470325, "grad_norm": 0.2097434252500534, "learning_rate": 1.657773963788245e-07, "loss": 0.0226, "step": 156060 }, { "epoch": 3.1187802246113265, "grad_norm": 0.14251694083213806, "learning_rate": 1.649376299001637e-07, "loss": 0.0259, "step": 156070 }, { "epoch": 3.1189800567523283, "grad_norm": 0.13440056145191193, "learning_rate": 1.6409999228381156e-07, "loss": 0.0202, "step": 156080 }, { "epoch": 3.1191798888933295, "grad_norm": 0.17913717031478882, "learning_rate": 1.6326448356556168e-07, "loss": 0.0239, "step": 156090 }, { "epoch": 3.1193797210343313, "grad_norm": 0.16779935359954834, "learning_rate": 1.6243110378110216e-07, "loss": 0.0232, "step": 156100 }, { "epoch": 3.1195795531753325, "grad_norm": 0.12483415752649307, "learning_rate": 1.6159985296602677e-07, "loss": 0.0273, "step": 156110 }, { "epoch": 3.1197793853163343, "grad_norm": 0.17908050119876862, "learning_rate": 1.6077073115584594e-07, "loss": 0.0252, "step": 156120 }, { "epoch": 3.119979217457336, "grad_norm": 0.18468047678470612, "learning_rate": 1.599437383859925e-07, "loss": 0.0238, "step": 156130 }, { "epoch": 3.1201790495983373, "grad_norm": 0.07004337012767792, "learning_rate": 1.591188746917771e-07, "loss": 0.0188, "step": 156140 }, { "epoch": 3.120378881739339, "grad_norm": 0.08651664853096008, "learning_rate": 1.5829614010844373e-07, "loss": 0.0206, "step": 156150 }, { "epoch": 3.1205787138803407, "grad_norm": 0.12711432576179504, "learning_rate": 1.5747553467113653e-07, "loss": 0.0314, "step": 156160 }, { "epoch": 3.120778546021342, "grad_norm": 0.17093433439731598, "learning_rate": 1.5665705841491075e-07, "loss": 0.0223, "step": 156170 }, { "epoch": 3.1209783781623437, "grad_norm": 0.12718500196933746, "learning_rate": 1.558407113747329e-07, "loss": 0.0252, "step": 156180 }, { "epoch": 3.121178210303345, "grad_norm": 0.09892411530017853, "learning_rate": 1.550264935854695e-07, "loss": 0.0175, "step": 156190 }, { "epoch": 3.1213780424443467, "grad_norm": 0.07660236209630966, "learning_rate": 1.5421440508190943e-07, "loss": 0.023, "step": 156200 }, { "epoch": 3.1215778745853484, "grad_norm": 0.10864216834306717, "learning_rate": 1.5340444589873604e-07, "loss": 0.0205, "step": 156210 }, { "epoch": 3.1217777067263497, "grad_norm": 0.1292649805545807, "learning_rate": 1.5259661607055497e-07, "loss": 0.0205, "step": 156220 }, { "epoch": 3.1219775388673514, "grad_norm": 0.1330270618200302, "learning_rate": 1.5179091563187752e-07, "loss": 0.0273, "step": 156230 }, { "epoch": 3.122177371008353, "grad_norm": 0.1095593124628067, "learning_rate": 1.5098734461711504e-07, "loss": 0.0216, "step": 156240 }, { "epoch": 3.1223772031493544, "grad_norm": 0.1287059783935547, "learning_rate": 1.501859030606012e-07, "loss": 0.0266, "step": 156250 }, { "epoch": 3.122577035290356, "grad_norm": 0.11057629436254501, "learning_rate": 1.4938659099656415e-07, "loss": 0.0233, "step": 156260 }, { "epoch": 3.122776867431358, "grad_norm": 0.17943419516086578, "learning_rate": 1.4858940845915436e-07, "loss": 0.019, "step": 156270 }, { "epoch": 3.122976699572359, "grad_norm": 0.10085058212280273, "learning_rate": 1.4779435548242793e-07, "loss": 0.0266, "step": 156280 }, { "epoch": 3.123176531713361, "grad_norm": 0.12526804208755493, "learning_rate": 1.4700143210034102e-07, "loss": 0.0213, "step": 156290 }, { "epoch": 3.123376363854362, "grad_norm": 0.14229334890842438, "learning_rate": 1.4621063834677205e-07, "loss": 0.0196, "step": 156300 }, { "epoch": 3.123576195995364, "grad_norm": 0.11809223890304565, "learning_rate": 1.4542197425549963e-07, "loss": 0.0224, "step": 156310 }, { "epoch": 3.1237760281363656, "grad_norm": 0.20646868646144867, "learning_rate": 1.4463543986021344e-07, "loss": 0.0193, "step": 156320 }, { "epoch": 3.123975860277367, "grad_norm": 0.13773204386234283, "learning_rate": 1.4385103519451993e-07, "loss": 0.0248, "step": 156330 }, { "epoch": 3.1241756924183686, "grad_norm": 0.08735430985689163, "learning_rate": 1.4306876029191452e-07, "loss": 0.0264, "step": 156340 }, { "epoch": 3.1243755245593703, "grad_norm": 0.14611873030662537, "learning_rate": 1.4228861518582602e-07, "loss": 0.0192, "step": 156350 }, { "epoch": 3.1245753567003716, "grad_norm": 0.10922999680042267, "learning_rate": 1.4151059990957228e-07, "loss": 0.0197, "step": 156360 }, { "epoch": 3.1247751888413733, "grad_norm": 0.1335422545671463, "learning_rate": 1.4073471449639886e-07, "loss": 0.0267, "step": 156370 }, { "epoch": 3.124975020982375, "grad_norm": 0.10519974678754807, "learning_rate": 1.399609589794404e-07, "loss": 0.0182, "step": 156380 }, { "epoch": 3.1251748531233763, "grad_norm": 0.15096814930438995, "learning_rate": 1.3918933339175378e-07, "loss": 0.0251, "step": 156390 }, { "epoch": 3.125374685264378, "grad_norm": 0.1421179622411728, "learning_rate": 1.3841983776630153e-07, "loss": 0.0189, "step": 156400 }, { "epoch": 3.1255745174053793, "grad_norm": 0.10354628413915634, "learning_rate": 1.3765247213595732e-07, "loss": 0.0231, "step": 156410 }, { "epoch": 3.125774349546381, "grad_norm": 0.08747539669275284, "learning_rate": 1.36887236533495e-07, "loss": 0.0187, "step": 156420 }, { "epoch": 3.1259741816873827, "grad_norm": 0.1718997210264206, "learning_rate": 1.361241309916106e-07, "loss": 0.025, "step": 156430 }, { "epoch": 3.126174013828384, "grad_norm": 0.11485272645950317, "learning_rate": 1.3536315554290023e-07, "loss": 0.024, "step": 156440 }, { "epoch": 3.1263738459693857, "grad_norm": 0.11698156595230103, "learning_rate": 1.346043102198713e-07, "loss": 0.0253, "step": 156450 }, { "epoch": 3.1265736781103874, "grad_norm": 0.10120584070682526, "learning_rate": 1.3384759505494227e-07, "loss": 0.0248, "step": 156460 }, { "epoch": 3.1267735102513887, "grad_norm": 0.08476205170154572, "learning_rate": 1.330930100804373e-07, "loss": 0.0205, "step": 156470 }, { "epoch": 3.1269733423923904, "grad_norm": 0.09088809043169022, "learning_rate": 1.323405553285917e-07, "loss": 0.0256, "step": 156480 }, { "epoch": 3.127173174533392, "grad_norm": 0.11451683193445206, "learning_rate": 1.3159023083154086e-07, "loss": 0.017, "step": 156490 }, { "epoch": 3.1273730066743934, "grad_norm": 0.12390744686126709, "learning_rate": 1.3084203662135364e-07, "loss": 0.0216, "step": 156500 }, { "epoch": 3.127572838815395, "grad_norm": 0.09658835083246231, "learning_rate": 1.3009597272997665e-07, "loss": 0.0236, "step": 156510 }, { "epoch": 3.1277726709563964, "grad_norm": 0.27313071489334106, "learning_rate": 1.293520391892844e-07, "loss": 0.0284, "step": 156520 }, { "epoch": 3.127972503097398, "grad_norm": 0.09977941960096359, "learning_rate": 1.286102360310626e-07, "loss": 0.0176, "step": 156530 }, { "epoch": 3.1281723352384, "grad_norm": 0.08897282183170319, "learning_rate": 1.2787056328699144e-07, "loss": 0.0231, "step": 156540 }, { "epoch": 3.128372167379401, "grad_norm": 0.14898426830768585, "learning_rate": 1.2713302098867342e-07, "loss": 0.0262, "step": 156550 }, { "epoch": 3.128571999520403, "grad_norm": 0.11570888757705688, "learning_rate": 1.2639760916761112e-07, "loss": 0.0225, "step": 156560 }, { "epoch": 3.1287718316614046, "grad_norm": 0.1530279815196991, "learning_rate": 1.2566432785522386e-07, "loss": 0.0252, "step": 156570 }, { "epoch": 3.128971663802406, "grad_norm": 0.1317049115896225, "learning_rate": 1.2493317708283104e-07, "loss": 0.0259, "step": 156580 }, { "epoch": 3.1291714959434076, "grad_norm": 0.0997820720076561, "learning_rate": 1.242041568816743e-07, "loss": 0.0199, "step": 156590 }, { "epoch": 3.1293713280844093, "grad_norm": 0.11416710913181305, "learning_rate": 1.234772672828899e-07, "loss": 0.0224, "step": 156600 }, { "epoch": 3.1295711602254106, "grad_norm": 0.11738868057727814, "learning_rate": 1.2275250831752515e-07, "loss": 0.0223, "step": 156610 }, { "epoch": 3.1297709923664123, "grad_norm": 0.10668708384037018, "learning_rate": 1.220298800165498e-07, "loss": 0.0242, "step": 156620 }, { "epoch": 3.1299708245074136, "grad_norm": 0.1068662777543068, "learning_rate": 1.2130938241083356e-07, "loss": 0.0197, "step": 156630 }, { "epoch": 3.1301706566484153, "grad_norm": 0.12900184094905853, "learning_rate": 1.2059101553114628e-07, "loss": 0.0269, "step": 156640 }, { "epoch": 3.130370488789417, "grad_norm": 0.13798263669013977, "learning_rate": 1.1987477940818004e-07, "loss": 0.0287, "step": 156650 }, { "epoch": 3.1305703209304183, "grad_norm": 0.10242124646902084, "learning_rate": 1.1916067407253262e-07, "loss": 0.0274, "step": 156660 }, { "epoch": 3.13077015307142, "grad_norm": 0.11735601723194122, "learning_rate": 1.1844869955470184e-07, "loss": 0.0216, "step": 156670 }, { "epoch": 3.1309699852124218, "grad_norm": 0.13433773815631866, "learning_rate": 1.1773885588511335e-07, "loss": 0.0241, "step": 156680 }, { "epoch": 3.131169817353423, "grad_norm": 0.10251256823539734, "learning_rate": 1.1703114309407626e-07, "loss": 0.0266, "step": 156690 }, { "epoch": 3.1313696494944248, "grad_norm": 0.14147083461284637, "learning_rate": 1.1632556121183857e-07, "loss": 0.0215, "step": 156700 }, { "epoch": 3.1315694816354265, "grad_norm": 0.1503414362668991, "learning_rate": 1.1562211026853176e-07, "loss": 0.0265, "step": 156710 }, { "epoch": 3.1317693137764278, "grad_norm": 0.1383795142173767, "learning_rate": 1.14920790294204e-07, "loss": 0.026, "step": 156720 }, { "epoch": 3.1319691459174295, "grad_norm": 0.15407821536064148, "learning_rate": 1.1422160131882021e-07, "loss": 0.0304, "step": 156730 }, { "epoch": 3.1321689780584308, "grad_norm": 0.10706286132335663, "learning_rate": 1.1352454337225093e-07, "loss": 0.0155, "step": 156740 }, { "epoch": 3.1323688101994325, "grad_norm": 0.08204164355993271, "learning_rate": 1.1282961648426682e-07, "loss": 0.025, "step": 156750 }, { "epoch": 3.132568642340434, "grad_norm": 0.11385013163089752, "learning_rate": 1.1213682068454968e-07, "loss": 0.0251, "step": 156760 }, { "epoch": 3.1327684744814355, "grad_norm": 0.08922059834003448, "learning_rate": 1.1144615600270913e-07, "loss": 0.0194, "step": 156770 }, { "epoch": 3.132968306622437, "grad_norm": 0.0968402847647667, "learning_rate": 1.1075762246823829e-07, "loss": 0.0234, "step": 156780 }, { "epoch": 3.133168138763439, "grad_norm": 0.09198605269193649, "learning_rate": 1.1007122011054693e-07, "loss": 0.0247, "step": 156790 }, { "epoch": 3.13336797090444, "grad_norm": 0.12142712622880936, "learning_rate": 1.0938694895896717e-07, "loss": 0.0248, "step": 156800 }, { "epoch": 3.133567803045442, "grad_norm": 0.1794610470533371, "learning_rate": 1.0870480904272561e-07, "loss": 0.0213, "step": 156810 }, { "epoch": 3.133767635186443, "grad_norm": 0.10490353405475616, "learning_rate": 1.0802480039096008e-07, "loss": 0.0274, "step": 156820 }, { "epoch": 3.133967467327445, "grad_norm": 0.11630386859178543, "learning_rate": 1.0734692303271953e-07, "loss": 0.0268, "step": 156830 }, { "epoch": 3.1341672994684466, "grad_norm": 0.11471199244260788, "learning_rate": 1.0667117699696416e-07, "loss": 0.0193, "step": 156840 }, { "epoch": 3.134367131609448, "grad_norm": 0.1256590038537979, "learning_rate": 1.0599756231255976e-07, "loss": 0.0276, "step": 156850 }, { "epoch": 3.1345669637504496, "grad_norm": 0.14555664360523224, "learning_rate": 1.0532607900827773e-07, "loss": 0.0221, "step": 156860 }, { "epoch": 3.1347667958914514, "grad_norm": 0.09136074781417847, "learning_rate": 1.0465672711281183e-07, "loss": 0.0224, "step": 156870 }, { "epoch": 3.1349666280324526, "grad_norm": 0.14437595009803772, "learning_rate": 1.0398950665474472e-07, "loss": 0.0219, "step": 156880 }, { "epoch": 3.1351664601734544, "grad_norm": 0.10933799296617508, "learning_rate": 1.0332441766258138e-07, "loss": 0.0278, "step": 156890 }, { "epoch": 3.1353662923144556, "grad_norm": 0.15829123556613922, "learning_rate": 1.0266146016474354e-07, "loss": 0.0189, "step": 156900 }, { "epoch": 3.1355661244554573, "grad_norm": 0.16182425618171692, "learning_rate": 1.0200063418954187e-07, "loss": 0.0307, "step": 156910 }, { "epoch": 3.135765956596459, "grad_norm": 0.15031538903713226, "learning_rate": 1.0134193976520378e-07, "loss": 0.0222, "step": 156920 }, { "epoch": 3.1359657887374603, "grad_norm": 0.14874687790870667, "learning_rate": 1.0068537691987346e-07, "loss": 0.0233, "step": 156930 }, { "epoch": 3.136165620878462, "grad_norm": 0.10346081107854843, "learning_rate": 1.000309456815951e-07, "loss": 0.0212, "step": 156940 }, { "epoch": 3.136365453019464, "grad_norm": 0.09985040128231049, "learning_rate": 9.93786460783297e-08, "loss": 0.0204, "step": 156950 }, { "epoch": 3.136565285160465, "grad_norm": 0.11190491914749146, "learning_rate": 9.872847813793273e-08, "loss": 0.0235, "step": 156960 }, { "epoch": 3.136765117301467, "grad_norm": 0.10306909680366516, "learning_rate": 9.808044188818755e-08, "loss": 0.021, "step": 156970 }, { "epoch": 3.1369649494424685, "grad_norm": 0.1242770254611969, "learning_rate": 9.7434537356772e-08, "loss": 0.0197, "step": 156980 }, { "epoch": 3.13716478158347, "grad_norm": 0.0952637642621994, "learning_rate": 9.679076457127511e-08, "loss": 0.0295, "step": 156990 }, { "epoch": 3.1373646137244715, "grad_norm": 0.10649337619543076, "learning_rate": 9.614912355920825e-08, "loss": 0.0219, "step": 157000 }, { "epoch": 3.137564445865473, "grad_norm": 0.1431466042995453, "learning_rate": 9.550961434797167e-08, "loss": 0.0327, "step": 157010 }, { "epoch": 3.1377642780064745, "grad_norm": 0.08847719430923462, "learning_rate": 9.487223696488245e-08, "loss": 0.0284, "step": 157020 }, { "epoch": 3.1379641101474762, "grad_norm": 0.10916148871183395, "learning_rate": 9.42369914371799e-08, "loss": 0.0208, "step": 157030 }, { "epoch": 3.1381639422884775, "grad_norm": 0.14373135566711426, "learning_rate": 9.360387779198676e-08, "loss": 0.026, "step": 157040 }, { "epoch": 3.1383637744294792, "grad_norm": 0.09228561818599701, "learning_rate": 9.297289605635916e-08, "loss": 0.0156, "step": 157050 }, { "epoch": 3.138563606570481, "grad_norm": 0.10067988187074661, "learning_rate": 9.23440462572478e-08, "loss": 0.0298, "step": 157060 }, { "epoch": 3.1387634387114822, "grad_norm": 0.09495957940816879, "learning_rate": 9.171732842151448e-08, "loss": 0.0242, "step": 157070 }, { "epoch": 3.138963270852484, "grad_norm": 0.10149134695529938, "learning_rate": 9.109274257593225e-08, "loss": 0.03, "step": 157080 }, { "epoch": 3.1391631029934857, "grad_norm": 0.1085602417588234, "learning_rate": 9.047028874717978e-08, "loss": 0.0262, "step": 157090 }, { "epoch": 3.139362935134487, "grad_norm": 0.16639208793640137, "learning_rate": 8.984996696185244e-08, "loss": 0.0193, "step": 157100 }, { "epoch": 3.1395627672754887, "grad_norm": 0.15025843679904938, "learning_rate": 8.923177724644572e-08, "loss": 0.0273, "step": 157110 }, { "epoch": 3.13976259941649, "grad_norm": 0.11469843983650208, "learning_rate": 8.86157196273718e-08, "loss": 0.0196, "step": 157120 }, { "epoch": 3.1399624315574917, "grad_norm": 0.11480497568845749, "learning_rate": 8.8001794130943e-08, "loss": 0.0204, "step": 157130 }, { "epoch": 3.1401622636984934, "grad_norm": 0.14862889051437378, "learning_rate": 8.739000078338278e-08, "loss": 0.029, "step": 157140 }, { "epoch": 3.1403620958394947, "grad_norm": 0.06883783638477325, "learning_rate": 8.678033961083132e-08, "loss": 0.0164, "step": 157150 }, { "epoch": 3.1405619279804964, "grad_norm": 0.14500756561756134, "learning_rate": 8.617281063932891e-08, "loss": 0.0175, "step": 157160 }, { "epoch": 3.140761760121498, "grad_norm": 0.11490871757268906, "learning_rate": 8.556741389483258e-08, "loss": 0.0253, "step": 157170 }, { "epoch": 3.1409615922624994, "grad_norm": 0.14248773455619812, "learning_rate": 8.496414940319941e-08, "loss": 0.0224, "step": 157180 }, { "epoch": 3.141161424403501, "grad_norm": 0.10203813016414642, "learning_rate": 8.436301719020323e-08, "loss": 0.0174, "step": 157190 }, { "epoch": 3.141361256544503, "grad_norm": 0.10031998157501221, "learning_rate": 8.376401728151794e-08, "loss": 0.0182, "step": 157200 }, { "epoch": 3.141561088685504, "grad_norm": 0.18777677416801453, "learning_rate": 8.31671497027342e-08, "loss": 0.0351, "step": 157210 }, { "epoch": 3.141760920826506, "grad_norm": 0.13508592545986176, "learning_rate": 8.257241447934827e-08, "loss": 0.0308, "step": 157220 }, { "epoch": 3.141960752967507, "grad_norm": 0.11362604051828384, "learning_rate": 8.19798116367676e-08, "loss": 0.0283, "step": 157230 }, { "epoch": 3.142160585108509, "grad_norm": 0.0975356474518776, "learning_rate": 8.138934120031083e-08, "loss": 0.0236, "step": 157240 }, { "epoch": 3.1423604172495105, "grad_norm": 0.11034125089645386, "learning_rate": 8.080100319519667e-08, "loss": 0.0218, "step": 157250 }, { "epoch": 3.142560249390512, "grad_norm": 0.15978053212165833, "learning_rate": 8.021479764656059e-08, "loss": 0.0272, "step": 157260 }, { "epoch": 3.1427600815315135, "grad_norm": 0.17029663920402527, "learning_rate": 7.96307245794381e-08, "loss": 0.026, "step": 157270 }, { "epoch": 3.1429599136725153, "grad_norm": 0.10639084130525589, "learning_rate": 7.904878401878701e-08, "loss": 0.0224, "step": 157280 }, { "epoch": 3.1431597458135165, "grad_norm": 0.17646872997283936, "learning_rate": 7.846897598946523e-08, "loss": 0.0261, "step": 157290 }, { "epoch": 3.1433595779545183, "grad_norm": 0.09933849424123764, "learning_rate": 7.789130051624183e-08, "loss": 0.0242, "step": 157300 }, { "epoch": 3.14355941009552, "grad_norm": 0.09354805946350098, "learning_rate": 7.731575762378596e-08, "loss": 0.0234, "step": 157310 }, { "epoch": 3.1437592422365213, "grad_norm": 0.12069692462682724, "learning_rate": 7.67423473366946e-08, "loss": 0.0219, "step": 157320 }, { "epoch": 3.143959074377523, "grad_norm": 0.10286755114793777, "learning_rate": 7.617106967945931e-08, "loss": 0.0342, "step": 157330 }, { "epoch": 3.1441589065185243, "grad_norm": 0.08722290396690369, "learning_rate": 7.560192467648275e-08, "loss": 0.0215, "step": 157340 }, { "epoch": 3.144358738659526, "grad_norm": 0.09653844684362411, "learning_rate": 7.503491235207328e-08, "loss": 0.0248, "step": 157350 }, { "epoch": 3.1445585708005277, "grad_norm": 0.1646471917629242, "learning_rate": 7.44700327304615e-08, "loss": 0.0269, "step": 157360 }, { "epoch": 3.144758402941529, "grad_norm": 0.11416404694318771, "learning_rate": 7.390728583577256e-08, "loss": 0.0179, "step": 157370 }, { "epoch": 3.1449582350825307, "grad_norm": 0.11795321851968765, "learning_rate": 7.334667169204279e-08, "loss": 0.0257, "step": 157380 }, { "epoch": 3.1451580672235324, "grad_norm": 0.11752098053693771, "learning_rate": 7.278819032323081e-08, "loss": 0.0213, "step": 157390 }, { "epoch": 3.1453578993645337, "grad_norm": 0.10361719131469727, "learning_rate": 7.223184175318976e-08, "loss": 0.0273, "step": 157400 }, { "epoch": 3.1455577315055354, "grad_norm": 0.16287538409233093, "learning_rate": 7.16776260056784e-08, "loss": 0.0276, "step": 157410 }, { "epoch": 3.145757563646537, "grad_norm": 0.07277201861143112, "learning_rate": 7.11255431043778e-08, "loss": 0.0309, "step": 157420 }, { "epoch": 3.1459573957875384, "grad_norm": 0.10076098144054413, "learning_rate": 7.05755930728802e-08, "loss": 0.0185, "step": 157430 }, { "epoch": 3.14615722792854, "grad_norm": 0.08496996760368347, "learning_rate": 7.002777593466126e-08, "loss": 0.0256, "step": 157440 }, { "epoch": 3.1463570600695414, "grad_norm": 0.09660394489765167, "learning_rate": 6.948209171313558e-08, "loss": 0.0244, "step": 157450 }, { "epoch": 3.146556892210543, "grad_norm": 0.11183866858482361, "learning_rate": 6.893854043160674e-08, "loss": 0.0225, "step": 157460 }, { "epoch": 3.146756724351545, "grad_norm": 0.18025505542755127, "learning_rate": 6.839712211330063e-08, "loss": 0.0241, "step": 157470 }, { "epoch": 3.146956556492546, "grad_norm": 0.12264122813940048, "learning_rate": 6.785783678133761e-08, "loss": 0.0269, "step": 157480 }, { "epoch": 3.147156388633548, "grad_norm": 0.19723646342754364, "learning_rate": 6.732068445876038e-08, "loss": 0.0278, "step": 157490 }, { "epoch": 3.1473562207745496, "grad_norm": 0.11471079289913177, "learning_rate": 6.678566516851725e-08, "loss": 0.0241, "step": 157500 }, { "epoch": 3.147556052915551, "grad_norm": 0.10860752314329147, "learning_rate": 6.62527789334566e-08, "loss": 0.024, "step": 157510 }, { "epoch": 3.1477558850565526, "grad_norm": 0.10696127265691757, "learning_rate": 6.572202577634912e-08, "loss": 0.0244, "step": 157520 }, { "epoch": 3.147955717197554, "grad_norm": 0.15115225315093994, "learning_rate": 6.519340571986554e-08, "loss": 0.0276, "step": 157530 }, { "epoch": 3.1481555493385556, "grad_norm": 0.11690046638250351, "learning_rate": 6.466691878658227e-08, "loss": 0.0269, "step": 157540 }, { "epoch": 3.1483553814795573, "grad_norm": 0.1384962797164917, "learning_rate": 6.414256499899795e-08, "loss": 0.0253, "step": 157550 }, { "epoch": 3.1485552136205586, "grad_norm": 0.16566084325313568, "learning_rate": 6.362034437950582e-08, "loss": 0.0272, "step": 157560 }, { "epoch": 3.1487550457615603, "grad_norm": 0.10732662677764893, "learning_rate": 6.310025695042132e-08, "loss": 0.026, "step": 157570 }, { "epoch": 3.148954877902562, "grad_norm": 0.16308604180812836, "learning_rate": 6.258230273394893e-08, "loss": 0.0252, "step": 157580 }, { "epoch": 3.1491547100435633, "grad_norm": 0.14433495700359344, "learning_rate": 6.206648175222651e-08, "loss": 0.0255, "step": 157590 }, { "epoch": 3.149354542184565, "grad_norm": 0.10905452072620392, "learning_rate": 6.155279402728641e-08, "loss": 0.0169, "step": 157600 }, { "epoch": 3.1495543743255667, "grad_norm": 0.1438639760017395, "learning_rate": 6.104123958107222e-08, "loss": 0.0269, "step": 157610 }, { "epoch": 3.149754206466568, "grad_norm": 0.15391460061073303, "learning_rate": 6.053181843542755e-08, "loss": 0.0252, "step": 157620 }, { "epoch": 3.1499540386075697, "grad_norm": 0.11330043524503708, "learning_rate": 6.002453061212943e-08, "loss": 0.0264, "step": 157630 }, { "epoch": 3.150153870748571, "grad_norm": 0.10486079752445221, "learning_rate": 5.951937613283831e-08, "loss": 0.028, "step": 157640 }, { "epoch": 3.1503537028895727, "grad_norm": 0.11154758185148239, "learning_rate": 5.901635501913694e-08, "loss": 0.025, "step": 157650 }, { "epoch": 3.1505535350305744, "grad_norm": 0.1297464519739151, "learning_rate": 5.851546729251367e-08, "loss": 0.0326, "step": 157660 }, { "epoch": 3.1507533671715757, "grad_norm": 0.08494112640619278, "learning_rate": 5.801671297436251e-08, "loss": 0.0252, "step": 157670 }, { "epoch": 3.1509531993125774, "grad_norm": 0.17245040833950043, "learning_rate": 5.752009208598863e-08, "loss": 0.0254, "step": 157680 }, { "epoch": 3.151153031453579, "grad_norm": 0.11423520743846893, "learning_rate": 5.70256046486084e-08, "loss": 0.0223, "step": 157690 }, { "epoch": 3.1513528635945804, "grad_norm": 0.10106977075338364, "learning_rate": 5.653325068334936e-08, "loss": 0.0237, "step": 157700 }, { "epoch": 3.151552695735582, "grad_norm": 0.11532524973154068, "learning_rate": 5.604303021124469e-08, "loss": 0.0214, "step": 157710 }, { "epoch": 3.1517525278765834, "grad_norm": 0.12541693449020386, "learning_rate": 5.555494325322763e-08, "loss": 0.0207, "step": 157720 }, { "epoch": 3.151952360017585, "grad_norm": 0.13283996284008026, "learning_rate": 5.506898983014819e-08, "loss": 0.0277, "step": 157730 }, { "epoch": 3.152152192158587, "grad_norm": 0.09822536259889603, "learning_rate": 5.458516996277308e-08, "loss": 0.0256, "step": 157740 }, { "epoch": 3.152352024299588, "grad_norm": 0.13344806432724, "learning_rate": 5.4103483671769095e-08, "loss": 0.0165, "step": 157750 }, { "epoch": 3.15255185644059, "grad_norm": 0.1256827414035797, "learning_rate": 5.362393097770868e-08, "loss": 0.0221, "step": 157760 }, { "epoch": 3.1527516885815916, "grad_norm": 0.1045474112033844, "learning_rate": 5.3146511901081e-08, "loss": 0.0293, "step": 157770 }, { "epoch": 3.152951520722593, "grad_norm": 0.10594908148050308, "learning_rate": 5.267122646227529e-08, "loss": 0.025, "step": 157780 }, { "epoch": 3.1531513528635946, "grad_norm": 0.07527274638414383, "learning_rate": 5.219807468160309e-08, "loss": 0.0237, "step": 157790 }, { "epoch": 3.1533511850045963, "grad_norm": 0.10935875028371811, "learning_rate": 5.172705657927046e-08, "loss": 0.0297, "step": 157800 }, { "epoch": 3.1535510171455976, "grad_norm": 0.09504035860300064, "learning_rate": 5.1258172175400185e-08, "loss": 0.0247, "step": 157810 }, { "epoch": 3.1537508492865993, "grad_norm": 0.10000553727149963, "learning_rate": 5.0791421490020696e-08, "loss": 0.0303, "step": 157820 }, { "epoch": 3.1539506814276006, "grad_norm": 0.113804891705513, "learning_rate": 5.032680454307714e-08, "loss": 0.0206, "step": 157830 }, { "epoch": 3.1541505135686023, "grad_norm": 0.10114531219005585, "learning_rate": 4.986432135440366e-08, "loss": 0.0218, "step": 157840 }, { "epoch": 3.154350345709604, "grad_norm": 0.09657087922096252, "learning_rate": 4.940397194377333e-08, "loss": 0.0222, "step": 157850 }, { "epoch": 3.1545501778506053, "grad_norm": 0.10900358855724335, "learning_rate": 4.894575633083709e-08, "loss": 0.023, "step": 157860 }, { "epoch": 3.154750009991607, "grad_norm": 0.10228817164897919, "learning_rate": 4.8489674535179273e-08, "loss": 0.0213, "step": 157870 }, { "epoch": 3.1549498421326088, "grad_norm": 0.10716626048088074, "learning_rate": 4.80357265762732e-08, "loss": 0.023, "step": 157880 }, { "epoch": 3.15514967427361, "grad_norm": 0.11306630820035934, "learning_rate": 4.758391247352001e-08, "loss": 0.0244, "step": 157890 }, { "epoch": 3.1553495064146118, "grad_norm": 0.140616312623024, "learning_rate": 4.7134232246220935e-08, "loss": 0.0255, "step": 157900 }, { "epoch": 3.1555493385556135, "grad_norm": 0.108499675989151, "learning_rate": 4.668668591357173e-08, "loss": 0.0226, "step": 157910 }, { "epoch": 3.1557491706966148, "grad_norm": 0.09468477964401245, "learning_rate": 4.624127349470708e-08, "loss": 0.0195, "step": 157920 }, { "epoch": 3.1559490028376165, "grad_norm": 0.11879013478755951, "learning_rate": 4.579799500864512e-08, "loss": 0.0212, "step": 157930 }, { "epoch": 3.1561488349786178, "grad_norm": 0.12171712517738342, "learning_rate": 4.5356850474326253e-08, "loss": 0.0206, "step": 157940 }, { "epoch": 3.1563486671196195, "grad_norm": 0.11710255593061447, "learning_rate": 4.4917839910590955e-08, "loss": 0.017, "step": 157950 }, { "epoch": 3.156548499260621, "grad_norm": 0.10708405822515488, "learning_rate": 4.4480963336190895e-08, "loss": 0.02, "step": 157960 }, { "epoch": 3.1567483314016225, "grad_norm": 0.09839652478694916, "learning_rate": 4.4046220769800027e-08, "loss": 0.0207, "step": 157970 }, { "epoch": 3.156948163542624, "grad_norm": 0.07987213134765625, "learning_rate": 4.361361222997573e-08, "loss": 0.0185, "step": 157980 }, { "epoch": 3.157147995683626, "grad_norm": 0.08603862673044205, "learning_rate": 4.318313773520877e-08, "loss": 0.0208, "step": 157990 }, { "epoch": 3.157347827824627, "grad_norm": 0.09687741845846176, "learning_rate": 4.275479730388443e-08, "loss": 0.0196, "step": 158000 }, { "epoch": 3.157547659965629, "grad_norm": 0.0913512110710144, "learning_rate": 4.232859095430475e-08, "loss": 0.0243, "step": 158010 }, { "epoch": 3.1577474921066306, "grad_norm": 0.1450098603963852, "learning_rate": 4.190451870466627e-08, "loss": 0.0308, "step": 158020 }, { "epoch": 3.157947324247632, "grad_norm": 0.10331729799509048, "learning_rate": 4.14825805730934e-08, "loss": 0.0319, "step": 158030 }, { "epoch": 3.1581471563886336, "grad_norm": 0.11724024266004562, "learning_rate": 4.1062776577610597e-08, "loss": 0.0188, "step": 158040 }, { "epoch": 3.158346988529635, "grad_norm": 0.1381962150335312, "learning_rate": 4.0645106736142416e-08, "loss": 0.0206, "step": 158050 }, { "epoch": 3.1585468206706366, "grad_norm": 0.13605855405330658, "learning_rate": 4.0229571066541237e-08, "loss": 0.0209, "step": 158060 }, { "epoch": 3.1587466528116384, "grad_norm": 0.08794087916612625, "learning_rate": 3.9816169586553986e-08, "loss": 0.018, "step": 158070 }, { "epoch": 3.1589464849526396, "grad_norm": 0.11386439204216003, "learning_rate": 3.940490231383875e-08, "loss": 0.019, "step": 158080 }, { "epoch": 3.1591463170936414, "grad_norm": 0.09499551355838776, "learning_rate": 3.899576926596482e-08, "loss": 0.0237, "step": 158090 }, { "epoch": 3.159346149234643, "grad_norm": 0.09592325240373611, "learning_rate": 3.858877046041265e-08, "loss": 0.0207, "step": 158100 }, { "epoch": 3.1595459813756444, "grad_norm": 0.12327367067337036, "learning_rate": 3.818390591456278e-08, "loss": 0.0202, "step": 158110 }, { "epoch": 3.159745813516646, "grad_norm": 0.19731292128562927, "learning_rate": 3.7781175645718037e-08, "loss": 0.0279, "step": 158120 }, { "epoch": 3.159945645657648, "grad_norm": 0.09013529121875763, "learning_rate": 3.738057967107578e-08, "loss": 0.0243, "step": 158130 }, { "epoch": 3.160145477798649, "grad_norm": 0.09278129786252975, "learning_rate": 3.698211800775009e-08, "loss": 0.0304, "step": 158140 }, { "epoch": 3.160345309939651, "grad_norm": 0.12871213257312775, "learning_rate": 3.658579067276069e-08, "loss": 0.0296, "step": 158150 }, { "epoch": 3.160545142080652, "grad_norm": 0.10572241991758347, "learning_rate": 3.619159768304403e-08, "loss": 0.0281, "step": 158160 }, { "epoch": 3.160744974221654, "grad_norm": 0.10780893266201019, "learning_rate": 3.579953905543665e-08, "loss": 0.0193, "step": 158170 }, { "epoch": 3.1609448063626555, "grad_norm": 0.11542366445064545, "learning_rate": 3.5409614806686253e-08, "loss": 0.0194, "step": 158180 }, { "epoch": 3.161144638503657, "grad_norm": 0.10650907456874847, "learning_rate": 3.502182495344619e-08, "loss": 0.0212, "step": 158190 }, { "epoch": 3.1613444706446585, "grad_norm": 0.13938049972057343, "learning_rate": 3.463616951228654e-08, "loss": 0.0241, "step": 158200 }, { "epoch": 3.1615443027856602, "grad_norm": 0.09339825809001923, "learning_rate": 3.4252648499683016e-08, "loss": 0.0225, "step": 158210 }, { "epoch": 3.1617441349266615, "grad_norm": 0.11484761536121368, "learning_rate": 3.38712619320114e-08, "loss": 0.0267, "step": 158220 }, { "epoch": 3.1619439670676632, "grad_norm": 0.13199730217456818, "learning_rate": 3.349200982556977e-08, "loss": 0.0267, "step": 158230 }, { "epoch": 3.162143799208665, "grad_norm": 0.11392354965209961, "learning_rate": 3.311489219656183e-08, "loss": 0.0196, "step": 158240 }, { "epoch": 3.1623436313496662, "grad_norm": 0.11932744085788727, "learning_rate": 3.2739909061091366e-08, "loss": 0.0211, "step": 158250 }, { "epoch": 3.162543463490668, "grad_norm": 0.12535656988620758, "learning_rate": 3.23670604351789e-08, "loss": 0.0305, "step": 158260 }, { "epoch": 3.1627432956316692, "grad_norm": 0.09726916253566742, "learning_rate": 3.199634633475057e-08, "loss": 0.0407, "step": 158270 }, { "epoch": 3.162943127772671, "grad_norm": 0.10975340753793716, "learning_rate": 3.162776677564927e-08, "loss": 0.0253, "step": 158280 }, { "epoch": 3.1631429599136727, "grad_norm": 0.1342572420835495, "learning_rate": 3.126132177361241e-08, "loss": 0.0311, "step": 158290 }, { "epoch": 3.163342792054674, "grad_norm": 0.11630073189735413, "learning_rate": 3.0897011344294126e-08, "loss": 0.0152, "step": 158300 }, { "epoch": 3.1635426241956757, "grad_norm": 0.11263852566480637, "learning_rate": 3.0534835503259754e-08, "loss": 0.0249, "step": 158310 }, { "epoch": 3.1637424563366774, "grad_norm": 0.0913613960146904, "learning_rate": 3.017479426598024e-08, "loss": 0.0147, "step": 158320 }, { "epoch": 3.1639422884776787, "grad_norm": 0.220321923494339, "learning_rate": 2.9816887647837746e-08, "loss": 0.0212, "step": 158330 }, { "epoch": 3.1641421206186804, "grad_norm": 0.10525142401456833, "learning_rate": 2.946111566412002e-08, "loss": 0.0241, "step": 158340 }, { "epoch": 3.1643419527596817, "grad_norm": 0.12310025840997696, "learning_rate": 2.9107478330026027e-08, "loss": 0.0166, "step": 158350 }, { "epoch": 3.1645417849006834, "grad_norm": 0.14757293462753296, "learning_rate": 2.87559756606548e-08, "loss": 0.0207, "step": 158360 }, { "epoch": 3.164741617041685, "grad_norm": 0.08779096603393555, "learning_rate": 2.8406607671033204e-08, "loss": 0.0196, "step": 158370 }, { "epoch": 3.1649414491826864, "grad_norm": 0.15612244606018066, "learning_rate": 2.8059374376082635e-08, "loss": 0.0276, "step": 158380 }, { "epoch": 3.165141281323688, "grad_norm": 0.08727159351110458, "learning_rate": 2.771427579063013e-08, "loss": 0.0224, "step": 158390 }, { "epoch": 3.16534111346469, "grad_norm": 0.17708933353424072, "learning_rate": 2.7371311929419442e-08, "loss": 0.0296, "step": 158400 }, { "epoch": 3.165540945605691, "grad_norm": 0.11243823915719986, "learning_rate": 2.703048280711107e-08, "loss": 0.0259, "step": 158410 }, { "epoch": 3.165740777746693, "grad_norm": 0.11913253366947174, "learning_rate": 2.669178843824893e-08, "loss": 0.0208, "step": 158420 }, { "epoch": 3.165940609887694, "grad_norm": 0.13421103358268738, "learning_rate": 2.6355228837315894e-08, "loss": 0.0232, "step": 158430 }, { "epoch": 3.166140442028696, "grad_norm": 0.10333379358053207, "learning_rate": 2.6020804018683786e-08, "loss": 0.0261, "step": 158440 }, { "epoch": 3.1663402741696975, "grad_norm": 0.10977037250995636, "learning_rate": 2.5688513996630083e-08, "loss": 0.0278, "step": 158450 }, { "epoch": 3.166540106310699, "grad_norm": 0.10640563070774078, "learning_rate": 2.5358358785365633e-08, "loss": 0.0206, "step": 158460 }, { "epoch": 3.1667399384517005, "grad_norm": 0.11339201033115387, "learning_rate": 2.5030338398979168e-08, "loss": 0.0294, "step": 158470 }, { "epoch": 3.1669397705927023, "grad_norm": 0.08537520468235016, "learning_rate": 2.470445285149281e-08, "loss": 0.0261, "step": 158480 }, { "epoch": 3.1671396027337035, "grad_norm": 0.12171056121587753, "learning_rate": 2.438070215682875e-08, "loss": 0.0186, "step": 158490 }, { "epoch": 3.1673394348747053, "grad_norm": 0.11947424709796906, "learning_rate": 2.4059086328809265e-08, "loss": 0.0168, "step": 158500 }, { "epoch": 3.167539267015707, "grad_norm": 0.09079667925834656, "learning_rate": 2.373960538117337e-08, "loss": 0.0226, "step": 158510 }, { "epoch": 3.1677390991567083, "grad_norm": 0.1277054101228714, "learning_rate": 2.34222593275768e-08, "loss": 0.0296, "step": 158520 }, { "epoch": 3.16793893129771, "grad_norm": 0.16258744895458221, "learning_rate": 2.3107048181569835e-08, "loss": 0.026, "step": 158530 }, { "epoch": 3.1681387634387113, "grad_norm": 0.1424972414970398, "learning_rate": 2.2793971956619475e-08, "loss": 0.0241, "step": 158540 }, { "epoch": 3.168338595579713, "grad_norm": 0.10983555763959885, "learning_rate": 2.248303066609836e-08, "loss": 0.024, "step": 158550 }, { "epoch": 3.1685384277207147, "grad_norm": 0.10821855813264847, "learning_rate": 2.2174224323290303e-08, "loss": 0.0247, "step": 158560 }, { "epoch": 3.168738259861716, "grad_norm": 0.09927038103342056, "learning_rate": 2.1867552941390314e-08, "loss": 0.0174, "step": 158570 }, { "epoch": 3.1689380920027177, "grad_norm": 0.11520066112279892, "learning_rate": 2.1563016533493462e-08, "loss": 0.0309, "step": 158580 }, { "epoch": 3.1691379241437194, "grad_norm": 0.16684959828853607, "learning_rate": 2.1260615112611572e-08, "loss": 0.0261, "step": 158590 }, { "epoch": 3.1693377562847207, "grad_norm": 0.08840832114219666, "learning_rate": 2.096034869166208e-08, "loss": 0.027, "step": 158600 }, { "epoch": 3.1695375884257224, "grad_norm": 0.15996180474758148, "learning_rate": 2.0662217283473618e-08, "loss": 0.0233, "step": 158610 }, { "epoch": 3.169737420566724, "grad_norm": 0.09987349063158035, "learning_rate": 2.0366220900780443e-08, "loss": 0.0279, "step": 158620 }, { "epoch": 3.1699372527077254, "grad_norm": 0.15269805490970612, "learning_rate": 2.0072359556222444e-08, "loss": 0.0245, "step": 158630 }, { "epoch": 3.170137084848727, "grad_norm": 0.12667812407016754, "learning_rate": 1.978063326236179e-08, "loss": 0.0251, "step": 158640 }, { "epoch": 3.1703369169897284, "grad_norm": 0.12864047288894653, "learning_rate": 1.9491042031655192e-08, "loss": 0.02, "step": 158650 }, { "epoch": 3.17053674913073, "grad_norm": 0.14244133234024048, "learning_rate": 1.920358587647053e-08, "loss": 0.0214, "step": 158660 }, { "epoch": 3.170736581271732, "grad_norm": 0.20292554795742035, "learning_rate": 1.8918264809092422e-08, "loss": 0.0251, "step": 158670 }, { "epoch": 3.170936413412733, "grad_norm": 0.14260433614253998, "learning_rate": 1.8635078841711118e-08, "loss": 0.018, "step": 158680 }, { "epoch": 3.171136245553735, "grad_norm": 0.15164931118488312, "learning_rate": 1.8354027986416942e-08, "loss": 0.0191, "step": 158690 }, { "epoch": 3.1713360776947366, "grad_norm": 0.12749843299388885, "learning_rate": 1.8075112255222516e-08, "loss": 0.0211, "step": 158700 }, { "epoch": 3.171535909835738, "grad_norm": 0.12230322510004044, "learning_rate": 1.7798331660040522e-08, "loss": 0.0268, "step": 158710 }, { "epoch": 3.1717357419767396, "grad_norm": 0.08520907908678055, "learning_rate": 1.7523686212689296e-08, "loss": 0.0213, "step": 158720 }, { "epoch": 3.1719355741177413, "grad_norm": 0.09486794471740723, "learning_rate": 1.7251175924909434e-08, "loss": 0.0327, "step": 158730 }, { "epoch": 3.1721354062587426, "grad_norm": 0.0980457291007042, "learning_rate": 1.6980800808336082e-08, "loss": 0.0228, "step": 158740 }, { "epoch": 3.1723352383997443, "grad_norm": 0.13716714084148407, "learning_rate": 1.671256087452111e-08, "loss": 0.0259, "step": 158750 }, { "epoch": 3.1725350705407456, "grad_norm": 0.09149454534053802, "learning_rate": 1.6446456134922016e-08, "loss": 0.0297, "step": 158760 }, { "epoch": 3.1727349026817473, "grad_norm": 0.09754329174757004, "learning_rate": 1.618248660090749e-08, "loss": 0.0275, "step": 158770 }, { "epoch": 3.172934734822749, "grad_norm": 0.09084473550319672, "learning_rate": 1.5920652283757386e-08, "loss": 0.027, "step": 158780 }, { "epoch": 3.1731345669637503, "grad_norm": 0.08129502832889557, "learning_rate": 1.566095319465166e-08, "loss": 0.0204, "step": 158790 }, { "epoch": 3.173334399104752, "grad_norm": 0.11100782454013824, "learning_rate": 1.540338934468144e-08, "loss": 0.0257, "step": 158800 }, { "epoch": 3.1735342312457537, "grad_norm": 0.12354519218206406, "learning_rate": 1.5147960744860136e-08, "loss": 0.0211, "step": 158810 }, { "epoch": 3.173734063386755, "grad_norm": 0.1514701545238495, "learning_rate": 1.4894667406090134e-08, "loss": 0.0295, "step": 158820 }, { "epoch": 3.1739338955277567, "grad_norm": 0.1380705088376999, "learning_rate": 1.4643509339196115e-08, "loss": 0.0243, "step": 158830 }, { "epoch": 3.1741337276687585, "grad_norm": 0.09650298953056335, "learning_rate": 1.439448655490283e-08, "loss": 0.0215, "step": 158840 }, { "epoch": 3.1743335598097597, "grad_norm": 0.08626259118318558, "learning_rate": 1.4147599063851769e-08, "loss": 0.0264, "step": 158850 }, { "epoch": 3.1745333919507615, "grad_norm": 0.1435304433107376, "learning_rate": 1.3902846876584497e-08, "loss": 0.0179, "step": 158860 }, { "epoch": 3.1747332240917627, "grad_norm": 0.1352832019329071, "learning_rate": 1.3660230003564867e-08, "loss": 0.022, "step": 158870 }, { "epoch": 3.1749330562327645, "grad_norm": 0.11648166924715042, "learning_rate": 1.341974845515126e-08, "loss": 0.0233, "step": 158880 }, { "epoch": 3.175132888373766, "grad_norm": 0.12543928623199463, "learning_rate": 1.318140224161879e-08, "loss": 0.0227, "step": 158890 }, { "epoch": 3.1753327205147674, "grad_norm": 0.14407262206077576, "learning_rate": 1.2945191373148203e-08, "loss": 0.0275, "step": 158900 }, { "epoch": 3.175532552655769, "grad_norm": 0.100104421377182, "learning_rate": 1.2711115859831424e-08, "loss": 0.0224, "step": 158910 }, { "epoch": 3.175732384796771, "grad_norm": 0.10373206436634064, "learning_rate": 1.2479175711660463e-08, "loss": 0.0247, "step": 158920 }, { "epoch": 3.175932216937772, "grad_norm": 0.11298806965351105, "learning_rate": 1.224937093855516e-08, "loss": 0.0234, "step": 158930 }, { "epoch": 3.176132049078774, "grad_norm": 0.1459295153617859, "learning_rate": 1.2021701550318787e-08, "loss": 0.0278, "step": 158940 }, { "epoch": 3.1763318812197756, "grad_norm": 0.27981826663017273, "learning_rate": 1.1796167556693549e-08, "loss": 0.021, "step": 158950 }, { "epoch": 3.176531713360777, "grad_norm": 0.10915923863649368, "learning_rate": 1.157276896729953e-08, "loss": 0.0294, "step": 158960 }, { "epoch": 3.1767315455017786, "grad_norm": 0.13398857414722443, "learning_rate": 1.1351505791684647e-08, "loss": 0.0255, "step": 158970 }, { "epoch": 3.17693137764278, "grad_norm": 0.12903966009616852, "learning_rate": 1.113237803930245e-08, "loss": 0.0236, "step": 158980 }, { "epoch": 3.1771312097837816, "grad_norm": 0.08026710152626038, "learning_rate": 1.0915385719512116e-08, "loss": 0.0238, "step": 158990 }, { "epoch": 3.1773310419247833, "grad_norm": 0.1352088898420334, "learning_rate": 1.0700528841584012e-08, "loss": 0.0171, "step": 159000 }, { "epoch": 3.1775308740657846, "grad_norm": 0.13777868449687958, "learning_rate": 1.048780741469968e-08, "loss": 0.0236, "step": 159010 }, { "epoch": 3.1777307062067863, "grad_norm": 0.11577289551496506, "learning_rate": 1.0277221447935192e-08, "loss": 0.0247, "step": 159020 }, { "epoch": 3.177930538347788, "grad_norm": 0.10943789780139923, "learning_rate": 1.006877095030001e-08, "loss": 0.0273, "step": 159030 }, { "epoch": 3.1781303704887893, "grad_norm": 0.08490467071533203, "learning_rate": 9.862455930692572e-09, "loss": 0.023, "step": 159040 }, { "epoch": 3.178330202629791, "grad_norm": 0.15317890048027039, "learning_rate": 9.658276397922494e-09, "loss": 0.0254, "step": 159050 }, { "epoch": 3.1785300347707928, "grad_norm": 0.14945171773433685, "learning_rate": 9.456232360716132e-09, "loss": 0.0284, "step": 159060 }, { "epoch": 3.178729866911794, "grad_norm": 0.13621722161769867, "learning_rate": 9.256323827705471e-09, "loss": 0.027, "step": 159070 }, { "epoch": 3.1789296990527958, "grad_norm": 0.12200632691383362, "learning_rate": 9.058550807428123e-09, "loss": 0.0219, "step": 159080 }, { "epoch": 3.179129531193797, "grad_norm": 0.1218276396393776, "learning_rate": 8.862913308332887e-09, "loss": 0.0174, "step": 159090 }, { "epoch": 3.1793293633347988, "grad_norm": 0.1324712336063385, "learning_rate": 8.669411338774192e-09, "loss": 0.0253, "step": 159100 }, { "epoch": 3.1795291954758005, "grad_norm": 0.11562694609165192, "learning_rate": 8.478044907028748e-09, "loss": 0.027, "step": 159110 }, { "epoch": 3.1797290276168018, "grad_norm": 0.08065866678953171, "learning_rate": 8.288814021256696e-09, "loss": 0.0208, "step": 159120 }, { "epoch": 3.1799288597578035, "grad_norm": 0.09250223636627197, "learning_rate": 8.101718689551562e-09, "loss": 0.0221, "step": 159130 }, { "epoch": 3.180128691898805, "grad_norm": 0.09451054781675339, "learning_rate": 7.916758919906952e-09, "loss": 0.0276, "step": 159140 }, { "epoch": 3.1803285240398065, "grad_norm": 0.12251333892345428, "learning_rate": 7.733934720216551e-09, "loss": 0.0209, "step": 159150 }, { "epoch": 3.180528356180808, "grad_norm": 0.14514964818954468, "learning_rate": 7.553246098290777e-09, "loss": 0.0354, "step": 159160 }, { "epoch": 3.1807281883218095, "grad_norm": 0.12466071546077728, "learning_rate": 7.374693061856786e-09, "loss": 0.0269, "step": 159170 }, { "epoch": 3.180928020462811, "grad_norm": 0.1477656215429306, "learning_rate": 7.198275618530703e-09, "loss": 0.0258, "step": 159180 }, { "epoch": 3.181127852603813, "grad_norm": 0.12043771147727966, "learning_rate": 7.0239937758564965e-09, "loss": 0.0175, "step": 159190 }, { "epoch": 3.181327684744814, "grad_norm": 0.08999805897474289, "learning_rate": 6.851847541278211e-09, "loss": 0.0279, "step": 159200 }, { "epoch": 3.181527516885816, "grad_norm": 0.10528189688920975, "learning_rate": 6.681836922151075e-09, "loss": 0.0244, "step": 159210 }, { "epoch": 3.1817273490268176, "grad_norm": 0.13800249993801117, "learning_rate": 6.513961925730394e-09, "loss": 0.0237, "step": 159220 }, { "epoch": 3.181927181167819, "grad_norm": 0.23578231036663055, "learning_rate": 6.348222559199313e-09, "loss": 0.019, "step": 159230 }, { "epoch": 3.1821270133088206, "grad_norm": 0.12451901286840439, "learning_rate": 6.1846188296244005e-09, "loss": 0.0232, "step": 159240 }, { "epoch": 3.182326845449822, "grad_norm": 0.0850919559597969, "learning_rate": 6.023150744005612e-09, "loss": 0.0305, "step": 159250 }, { "epoch": 3.1825266775908236, "grad_norm": 0.11674558371305466, "learning_rate": 5.863818309231884e-09, "loss": 0.0214, "step": 159260 }, { "epoch": 3.1827265097318254, "grad_norm": 0.16104091703891754, "learning_rate": 5.7066215321144315e-09, "loss": 0.0322, "step": 159270 }, { "epoch": 3.1829263418728266, "grad_norm": 0.12436621636152267, "learning_rate": 5.551560419370105e-09, "loss": 0.024, "step": 159280 }, { "epoch": 3.1831261740138284, "grad_norm": 0.09962650388479233, "learning_rate": 5.398634977615835e-09, "loss": 0.0253, "step": 159290 }, { "epoch": 3.18332600615483, "grad_norm": 0.12399187684059143, "learning_rate": 5.247845213390834e-09, "loss": 0.02, "step": 159300 }, { "epoch": 3.1835258382958314, "grad_norm": 0.09432385861873627, "learning_rate": 5.099191133134396e-09, "loss": 0.0226, "step": 159310 }, { "epoch": 3.183725670436833, "grad_norm": 0.13136164844036102, "learning_rate": 4.952672743202546e-09, "loss": 0.03, "step": 159320 }, { "epoch": 3.183925502577835, "grad_norm": 0.11105401813983917, "learning_rate": 4.808290049840292e-09, "loss": 0.0247, "step": 159330 }, { "epoch": 3.184125334718836, "grad_norm": 0.0820915699005127, "learning_rate": 4.666043059231573e-09, "loss": 0.0187, "step": 159340 }, { "epoch": 3.184325166859838, "grad_norm": 0.11116821318864822, "learning_rate": 4.525931777438208e-09, "loss": 0.0256, "step": 159350 }, { "epoch": 3.184524999000839, "grad_norm": 0.10509902238845825, "learning_rate": 4.387956210455402e-09, "loss": 0.0218, "step": 159360 }, { "epoch": 3.184724831141841, "grad_norm": 0.12358946353197098, "learning_rate": 4.252116364172887e-09, "loss": 0.0271, "step": 159370 }, { "epoch": 3.1849246632828425, "grad_norm": 0.11936016380786896, "learning_rate": 4.118412244397129e-09, "loss": 0.0288, "step": 159380 }, { "epoch": 3.185124495423844, "grad_norm": 0.08907594531774521, "learning_rate": 3.986843856834677e-09, "loss": 0.0292, "step": 159390 }, { "epoch": 3.1853243275648455, "grad_norm": 0.09686427563428879, "learning_rate": 3.8574112071143585e-09, "loss": 0.0222, "step": 159400 }, { "epoch": 3.1855241597058472, "grad_norm": 0.12516701221466064, "learning_rate": 3.7301143007539844e-09, "loss": 0.0212, "step": 159410 }, { "epoch": 3.1857239918468485, "grad_norm": 0.1273704171180725, "learning_rate": 3.6049531431991967e-09, "loss": 0.0248, "step": 159420 }, { "epoch": 3.1859238239878502, "grad_norm": 0.09315718710422516, "learning_rate": 3.4819277397957208e-09, "loss": 0.022, "step": 159430 }, { "epoch": 3.186123656128852, "grad_norm": 0.10767091810703278, "learning_rate": 3.3610380957949107e-09, "loss": 0.0192, "step": 159440 }, { "epoch": 3.1863234882698532, "grad_norm": 0.09546950459480286, "learning_rate": 3.242284216364855e-09, "loss": 0.022, "step": 159450 }, { "epoch": 3.186523320410855, "grad_norm": 0.10496115684509277, "learning_rate": 3.1256661065792726e-09, "loss": 0.0239, "step": 159460 }, { "epoch": 3.1867231525518562, "grad_norm": 0.10750192403793335, "learning_rate": 3.011183771411963e-09, "loss": 0.0248, "step": 159470 }, { "epoch": 3.186922984692858, "grad_norm": 0.11071279644966125, "learning_rate": 2.8988372157645606e-09, "loss": 0.0198, "step": 159480 }, { "epoch": 3.1871228168338597, "grad_norm": 0.15226614475250244, "learning_rate": 2.788626444433229e-09, "loss": 0.028, "step": 159490 }, { "epoch": 3.187322648974861, "grad_norm": 0.08549154549837112, "learning_rate": 2.6805514621197625e-09, "loss": 0.0295, "step": 159500 }, { "epoch": 3.1875224811158627, "grad_norm": 0.12545140087604523, "learning_rate": 2.57461227344824e-09, "loss": 0.0246, "step": 159510 }, { "epoch": 3.1877223132568644, "grad_norm": 0.07565226405858994, "learning_rate": 2.47080888293727e-09, "loss": 0.0241, "step": 159520 }, { "epoch": 3.1879221453978657, "grad_norm": 0.11934711039066315, "learning_rate": 2.3691412950277437e-09, "loss": 0.0194, "step": 159530 }, { "epoch": 3.1881219775388674, "grad_norm": 0.12676165997982025, "learning_rate": 2.2696095140606334e-09, "loss": 0.03, "step": 159540 }, { "epoch": 3.188321809679869, "grad_norm": 0.12576833367347717, "learning_rate": 2.172213544282542e-09, "loss": 0.0199, "step": 159550 }, { "epoch": 3.1885216418208704, "grad_norm": 0.12419548630714417, "learning_rate": 2.076953389862357e-09, "loss": 0.023, "step": 159560 }, { "epoch": 3.188721473961872, "grad_norm": 0.07138141989707947, "learning_rate": 1.983829054863495e-09, "loss": 0.0243, "step": 159570 }, { "epoch": 3.1889213061028734, "grad_norm": 0.1498715877532959, "learning_rate": 1.8928405432716567e-09, "loss": 0.0296, "step": 159580 }, { "epoch": 3.189121138243875, "grad_norm": 0.12691760063171387, "learning_rate": 1.8039878589615201e-09, "loss": 0.0202, "step": 159590 }, { "epoch": 3.189320970384877, "grad_norm": 0.09122638404369354, "learning_rate": 1.717271005735599e-09, "loss": 0.0253, "step": 159600 }, { "epoch": 3.189520802525878, "grad_norm": 0.14058344066143036, "learning_rate": 1.6326899873020384e-09, "loss": 0.0228, "step": 159610 }, { "epoch": 3.18972063466688, "grad_norm": 0.08443240821361542, "learning_rate": 1.5502448072690633e-09, "loss": 0.0311, "step": 159620 }, { "epoch": 3.1899204668078815, "grad_norm": 0.13959890604019165, "learning_rate": 1.4699354691616318e-09, "loss": 0.0195, "step": 159630 }, { "epoch": 3.190120298948883, "grad_norm": 0.12714506685733795, "learning_rate": 1.3917619764047818e-09, "loss": 0.0231, "step": 159640 }, { "epoch": 3.1903201310898845, "grad_norm": 0.0949905514717102, "learning_rate": 1.3157243323402846e-09, "loss": 0.0255, "step": 159650 }, { "epoch": 3.1905199632308863, "grad_norm": 0.13054583966732025, "learning_rate": 1.2418225402210937e-09, "loss": 0.0227, "step": 159660 }, { "epoch": 3.1907197953718875, "grad_norm": 0.1725926548242569, "learning_rate": 1.1700566032002424e-09, "loss": 0.0232, "step": 159670 }, { "epoch": 3.1909196275128893, "grad_norm": 0.1516382247209549, "learning_rate": 1.1004265243419464e-09, "loss": 0.0183, "step": 159680 }, { "epoch": 3.1911194596538905, "grad_norm": 0.1354469209909439, "learning_rate": 1.0329323066216035e-09, "loss": 0.0218, "step": 159690 }, { "epoch": 3.1913192917948923, "grad_norm": 0.13952016830444336, "learning_rate": 9.675739529257933e-10, "loss": 0.0167, "step": 159700 }, { "epoch": 3.191519123935894, "grad_norm": 0.1014646589756012, "learning_rate": 9.043514660411756e-10, "loss": 0.0232, "step": 159710 }, { "epoch": 3.1917189560768953, "grad_norm": 0.13333772122859955, "learning_rate": 8.432648486766948e-10, "loss": 0.03, "step": 159720 }, { "epoch": 3.191918788217897, "grad_norm": 0.1321529597043991, "learning_rate": 7.843141034302725e-10, "loss": 0.021, "step": 159730 }, { "epoch": 3.1921186203588987, "grad_norm": 0.12009993195533752, "learning_rate": 7.274992328276664e-10, "loss": 0.0156, "step": 159740 }, { "epoch": 3.1923184524999, "grad_norm": 0.07384607940912247, "learning_rate": 6.728202393002648e-10, "loss": 0.0184, "step": 159750 }, { "epoch": 3.1925182846409017, "grad_norm": 0.08189737796783447, "learning_rate": 6.202771251739847e-10, "loss": 0.0254, "step": 159760 }, { "epoch": 3.1927181167819034, "grad_norm": 0.10515767335891724, "learning_rate": 5.69869892697028e-10, "loss": 0.0241, "step": 159770 }, { "epoch": 3.1929179489229047, "grad_norm": 0.13255910575389862, "learning_rate": 5.215985440232274e-10, "loss": 0.0205, "step": 159780 }, { "epoch": 3.1931177810639064, "grad_norm": 0.11086281388998032, "learning_rate": 4.754630812175976e-10, "loss": 0.0224, "step": 159790 }, { "epoch": 3.1933176132049077, "grad_norm": 0.1004910096526146, "learning_rate": 4.314635062452332e-10, "loss": 0.0236, "step": 159800 }, { "epoch": 3.1935174453459094, "grad_norm": 0.1253267377614975, "learning_rate": 3.8959982099351366e-10, "loss": 0.0241, "step": 159810 }, { "epoch": 3.193717277486911, "grad_norm": 0.14591839909553528, "learning_rate": 3.4987202723879563e-10, "loss": 0.0251, "step": 159820 }, { "epoch": 3.1939171096279124, "grad_norm": 0.09884215891361237, "learning_rate": 3.1228012669082264e-10, "loss": 0.0153, "step": 159830 }, { "epoch": 3.194116941768914, "grad_norm": 0.07828488200902939, "learning_rate": 2.7682412094831577e-10, "loss": 0.0255, "step": 159840 }, { "epoch": 3.194316773909916, "grad_norm": 0.10005982965230942, "learning_rate": 2.4350401152672954e-10, "loss": 0.0319, "step": 159850 }, { "epoch": 3.194516606050917, "grad_norm": 0.10220637917518616, "learning_rate": 2.1231979984714934e-10, "loss": 0.0184, "step": 159860 }, { "epoch": 3.194716438191919, "grad_norm": 0.14904305338859558, "learning_rate": 1.8327148724739397e-10, "loss": 0.0291, "step": 159870 }, { "epoch": 3.1949162703329206, "grad_norm": 0.14672671258449554, "learning_rate": 1.5635907497091318e-10, "loss": 0.017, "step": 159880 }, { "epoch": 3.195116102473922, "grad_norm": 0.13232770562171936, "learning_rate": 1.3158256415568558e-10, "loss": 0.0287, "step": 159890 }, { "epoch": 3.1953159346149236, "grad_norm": 0.10750345885753632, "learning_rate": 1.0894195586752531e-10, "loss": 0.0201, "step": 159900 }, { "epoch": 3.195515766755925, "grad_norm": 0.13422970473766327, "learning_rate": 8.843725107232636e-11, "loss": 0.0309, "step": 159910 }, { "epoch": 3.1957155988969266, "grad_norm": 0.16035497188568115, "learning_rate": 7.006845064716493e-11, "loss": 0.0222, "step": 159920 }, { "epoch": 3.1959154310379283, "grad_norm": 0.11228277534246445, "learning_rate": 5.383555537474827e-11, "loss": 0.0296, "step": 159930 }, { "epoch": 3.1961152631789296, "grad_norm": 0.12831571698188782, "learning_rate": 3.973856595451686e-11, "loss": 0.0304, "step": 159940 }, { "epoch": 3.1963150953199313, "grad_norm": 0.1293741762638092, "learning_rate": 2.777748298044003e-11, "loss": 0.0164, "step": 159950 }, { "epoch": 3.196514927460933, "grad_norm": 0.15037615597248077, "learning_rate": 1.7952306963220368e-11, "loss": 0.0206, "step": 159960 }, { "epoch": 3.1967147596019343, "grad_norm": 0.14393801987171173, "learning_rate": 1.026303833029374e-11, "loss": 0.0236, "step": 159970 }, { "epoch": 3.196914591742936, "grad_norm": 0.21436871588230133, "learning_rate": 4.709677409175939e-12, "loss": 0.0295, "step": 159980 }, { "epoch": 3.1971144238839373, "grad_norm": 0.12945881485939026, "learning_rate": 1.292224427462685e-12, "loss": 0.0199, "step": 159990 }, { "epoch": 3.197314256024939, "grad_norm": 0.15255820751190186, "learning_rate": 1.0679540585201153e-14, "loss": 0.0342, "step": 160000 }, { "epoch": 3.197314256024939, "step": 160000, "total_flos": 0.0, "train_loss": 0.03381463914900087, "train_runtime": 96412.0494, "train_samples_per_second": 212.422, "train_steps_per_second": 1.66 } ], "logging_steps": 10, "max_steps": 160000, "num_input_tokens_seen": 0, "num_train_epochs": 4, "save_steps": 40000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 16, "trial_name": null, "trial_params": null }